Metalepsis: Twikent.

I was not familiar with the term metalepsis, which the OED (entry revised 2001) defines thus:

The rhetorical figure consisting in the metonymical substitution of one word for another which is itself a metonym; (more generally) any metaphorical usage resulting from a series or succession of figurative substitutions. Also: an instance of this.

It’s from post-classical Latin metalepsis < ancient Greek μετάληψις ‘alternation, succession.’ Some of the citations are quite lively:

1550 Transsumpcion, is when by degrees we go to yᵗ that is shewed as: he hyd hym selfe in the blacke dennes. By blacke, is vnder stand ful of darkenes & consequently stepe downe, and verye depe. [Margin] Metalepsis.
R. Sherry, Treatise of Schemes & Tropes sig. Cv

1589 The figure Metalepsis, which I call the farfet, as when we had rather fetch a word a great way off then to vse one nerer hand to expresse the matter aswel & plainer.
G. Puttenham, Arte of English Poesie iii. xvii. 152

1930 Naturally a man who could combine a synecdoche and a metalepsis with intent to deceive would be capable of anything.
Journal Royal Statist. Society vol. 93 237

1975 In a metalepsis a word is substituted metonymically for a word in a previous trope, so that a metalepsis can be called, maddeningly but accurately, a metonymy of a metonymy.
H. Bloom, Map of Misreading ii. v. 102

(Puttenham’s farfet = far-fetched.) But even livelier is the Icelandic example given in the Wikipedia article, introduced as follows:

The word twikent (twice-kenned) is used for once-removed metalepsis involving kennings. If a kenning has more than three elements, it is said to be rekit (“extended”). Kennings of up to seven elements are recorded in skaldic verse. Snorri Sturluson characterises five-element kennings as an acceptable license but cautions against more extreme constructions […]

nausta blakks hlé-mána gífrs drífu gim-slöngvir “fire-brandisher of blizzard of ogress of protection-moon of steed of boat-shed” — from Þórálfs drápa Skólmssonar, by Þórður Sjáreksson (this is the longest kenning found in skaldic poetry; it simply means “warrior”.)

Trevor Joyce, who sent it to me, called it “a bit of a head-scratcher,” and I confess I have a hard time parsing it; I think Snorri was right to warn of the peril of “more extreme constructions.”


  1. J.W. Brewer says

    I am skeptical that the last thing “simply” means warrior. “Non-consecutive-term-serving allegedly-bastard-spawning Buffalonian” (which might sound quite impressive in Old Norse) might uniquely *refer* to Grover Cleveland. But I don’t think it “simply means” Grover Cleveland, for useful values of “simply” and “means.”

  2. J.W. Brewer says

    Or, to take an example I tracked down on the internet and did not compose myself, you could perhaps describe “rootin,’ tootin,’ six-gun shootin’ ring tailed buckaroo” as “simply meaning” cowboy, but is that the most useful analysis?

  3. Jen in Edinburgh says

    This is all reminding me of the nonsense about the ink-lined plane, which I am not going to put any more effort into remembering when I should be in bed!

  4. Twikent is a Twi city in Ghanistan in Central Asia.

  5. Wiki’s English examples seem to be something of a much more common and tamer sort, just expansions of an existing metaphor (“rescued from the rabbit hole”). Quintilian disapprovingly gives a couple of Greek examples of the real thing, such as Chiron the centaur being called Hesson because his name is homonymous with χείρων “worse”, for which ἥσσων is a synonym.

  6. Wikipedia is a bit of a mess dealing with the rhetorical figure of speech and the narrative discourse technique. English just sort of runs them together. French has two separate pages. German introduces them as separate concepts and then covers each in turn.

  7. It just occurred to me to look up μετάληψη in my Modern Greek–English dictionary, where I learned that it means ‘[Holy] Communion, Eucharist, last rites.’

  8. The Journal of the Royal Statistical Society paper that the OED quotes, “The Vaccination Problem,” is, sadly enough, timely. The “man” in question is Edward Jenner and the trope is his coining of Variolae vaccinae for cowpox. But really the paper’s author (Major Greenwood) is mocking Charles Creighton rant that used those terms accusingly, and which is, of course, still regularly cited 133 years on as an authority by the wake-up-sheeple do-your-own-research internet crowd.

  9. Charles Creighton rant

    “A palpable catachresis”! Literally!

  10. I just came across a startling use of the word “vaccine” – Tennyson, in “Persia” (1827), writes

    “And further east, where, broadly rolled,
    Old Indus pours his stream of gold;
    And there, where tumbling deep and hoarse,
    Blue Ganga leaves her vaccine source; …”

    The pertinent note, in Ricks’ complete edition:

    54. 1827 note: “The cavern in the ridge of Himmalah, whence the Ganges seems to derive its original springs, has been moulded, by the mind of Hindoo superstition, into the head of a cow’. Paden shows that this was from Rennell. There was a discussion of Rennell and the cow’s-mouth in QR xvii (1817) 409. QR influenced other poems of 1827, and there were copies at Somersby (Lincoln). Cp. The Ganges (p. 162).

  11. Of course, both the cow and the Ganges are considered physical manifestations of divine motherhood in Hinduism.

  12. A few pieces of British slang of the last century are recognisably twikent: eg (to give someone) “the Spanish archer” –> “El Bow” –> to fire (from a job), break up with (a partner), etc.

    Contra JW Brewer, I think it’d be quite accurate to say that “the Spanish archer” does simply mean the termination of employment/relationship — the wordplay adds flavour to the telling, but doesn’t affect the meaning. Reading the skaldic kennings similarly seems plausible to me.

  13. The actual form the kenning occurs in is this:

    Ok gimsløngvir ganga
    gífrs hlémána drífu
    nausta blakks it næsta
    Norðmanna gram þorði.

    Already with the nominative sløngvir, a listener familiar with the tradition might guess that we have a warrior, especially in connection with gim: literally “fire”, but a common base-element for sword-kennings. A “sword-slinger” = “warrior”.

    Want we want is for “fire” to be modified by a word that associates it with war or killing or something, and we get that in the next verse: “troll-woman’s protection-moon’s blizzard’s”. The “blizzard” part will, again by strong convention, be naturally taken as a base element in a battle-kenning, which is already what we’re looking for: “the fire of some kind of blizzard” = “fire of battle” = “sword”. For “blizzard” to mean “battle”, it wants modification, probably with some kind of weaponry, which we’ve already had. If you see “moons”, they’re often shields; the “protection” part reinforces that. And some kind of foe, such as a giant or troll, of a shield will be an axe — the weapon-word we’ve been looking for to modify blizzard. This is actually presented in a rather friendly order, basically “the fire-slinger (already probably ‘sword-slinger’) of the trolls of protection moons’ (a nice unit readily taken as axes) blizzard”.

    Most of the important bits are already clear, but we still have nausta blakks in the following verse. This obviously means “ships” (“steeds of boathouses” should be pretty transparent even to someone who doesn’t know the tradition at all), in the genitive. It makes most sense to attach it to the “protection-moons”, both because this compound hasn’t itself been modified, and because the imagery of moon-shaped shields hanging on the sides of ships is a common image in kennings.

    This does feel like a rather padded kenning, and I’m not sure it’s nearly as interesting as some of the shorter but more imaginative ones, but I suppose its sheer length does deserve some credit. It’s maybe also worth pointing out that this kenning doesn’t have much in the way of mythological components to it (the gífr is a supernatural element, but not very religiously specific), despite its length. This is not unusual for eleventh-century poetry; skalds seem to have become shy of using kennings that directly alluded to pagan religion in that period. They don’t seem to have had any compunctions about memorizing and transmitting existing pagan poems, though, which is what allowed the later “revival” of pagan kennings by Christians who were a bit more secure in their religion (including by Snorri Sturluson, whose Prose Edda includes so much mythological explanation precisely because he felt it was important context for poets to know).

  14. Thanks very much for explicating the kenning so thoroughly!

  15. David Eddyshaw says

    I suppose the fairly common Kusaasi personal name Akudugu is twikent.

    It’s Akudug in Kusaal, where A is the particle that precedes (almost) all personal names, and kudug is “iron” or “nail” (replaced as a common noun now by the formally-plural form kut.)

    The “nail” in question is a nail driven into a tree to mark that the win “spiritual individuality” of the tree is the thus-named child’s sigir “guardian spirit.” (You can just be called “Atiga” Atiig “Tree”, but this is a more allusive way of making the same reference.)

    I had an extremely capable colleague called Akudibillah, Akudbil “Little Nail”; this name signifies that he shares the same tree-sigir as his elder brother Akudugu.

    I’ve previously mentioned the name “Anaba” Ana’ab “Chief”, which in fact is a birth-circumstance name given to the sole survivor of a pair of twins (“chief” means “afterbirth”, because a chief’s retainers precede him on leaving the house; “afterbirth” is given as an apotropaic name, to throw the malign spiritual forces that killed the other twin off the scent.)

  16. David Eddyshaw says

    This sort of very indirect allusion turns up often in Kusaal proverbs, too, to the degree that it’s often by no means obvious from the literal sense what the proverb actually means (this is a a feature rather than a bug: proverbs are often used for speaking truth to power or saying other potentially unwelcome things, so a layer of ostensible plausible deniability saves everyone’s blushes.)

    Ku’om zɔtnɛ bian’ar zug.
    “Water runs on mud.”

    Ah. Yes. Yes it does. And your point was …?
    This sort of thing seems to be very common throughout Africa, judging by Chapter 14 of

  17. Stu Clayton says

    And your point was …?

    Et alors ?? “Muddy waters don’t run deep” ? “Where there’s water there’s mud” ? “In for a penny, in for a pound” ? “Einmal gebockt ist nicht gelammt” ?

    (this is a feature rather than a bug)

    At work I call that a feature bug. Something stupid that the software makers nevertheless extol.

  18. From the earliest days of LH:

    “But one can only be a successful official when living in a peaceful residence,” she countered, using a Chinese saying.

    “I’d rather ride a donkey to look for my horse,” I replied with another.

    “I think you’re painting legs on a snake here.”

    “The only snake is being reflected in my soup.” This went on for awhile, but I knew when she started shifting to old Taiwanese sayings that I was fighting a losing battle. “I’m going to take a walk around the neighborbood,” I said firmly, and as she tried to figure out what I was really trying to say with this apparently unknown ancient saying, I took advantage of the lull to beat a hasty retreat before I was drowned in irrelevant flowery rhetoric.

  19. David Eddyshaw says


    “You scratch my back and I’ll scratch yours” in fact.

    A more perspicuous but perhaps related usage appears in an ancient proverbial saying I encountered while waiting my turn in a government agency office elsewhere in anglophone West Africa: “A car does not run without petrol.”

  20. very common throughout Africa, judging by Chapter 14

    It also mentions there their physical forms among the Akan, Fante flags and Ashanti gold weights. I don’t trust that we actually know the correct proverbs for the several of each we have here.

    Some anonymous user who geolocates to Bristol has vandalized the Asafo Flags Wikipedia page by inserting the words ohio and muse. I cannot imagine why.

  21. Stu Clayton says

    I just reverted that vandalism. [They call me The Equalizer.]

  22. David Eddyshaw says

    I wonder what it means by saying the flags are called “frank” in Fante? That’s not even a possible Fante word. Maybe they mean “in the local English, as spoken by Fante.”

    Actually, it might be just an inaccurate transcription of an actual Fante loan from English. (Fr- for fl- would make sense on that basis. Northern Ghanaian mockery of Akans’ English conventionally has them saying “probrem” for “problem.” Which they do … And lots of loans from English into Ghanaian languages seem to involve the locals hearing Brits as pronouncing the English vowels as nasalised, like Mampruli mantooka “motor car” and mankyɛs “matches.”

  23. I think it’s usually given as [ɔ]frankaa.

    Num 10:14:

    Na Yuda mma nsra ɔfrankaa no tui kan, sɛ wɔn dɔm ahorow te; na ne dɔm no so na Aminadab ba Nahason di

    I don’t know whether it’s a loanword.

  24. Stu Clayton says

    the first standard to set out, troop by troop, was the division of Judah. In command of its troops was Nahshon son of Amminadab

    It is very right that division should be standardized. Otherwise it’s all too easy to make rounding errors.

  25. David Eddyshaw says


    Thanks! Well found.

    Christaller’s wonderful dictionary (from 1881) actually has that, glossed as “flag, ensign, colours, banner, standard.” It doesn’t have any obvious intra-Akan etymology, and Christaller cites the Gã aflanga. That looks rather as if it ought to have been borrowed from Portuguese, though I can’t think of a plausible source word. Perhaps ultimately English after all.

    If anyone has the fortitude to edit WP, correcting “frank” to ɔfrankaa would be reasonable, I think.

  26. The form used in African Arts (going back to a paper by George Nelson Preston in 1975 and occasionally right up to the last volume in 2018) seems to be frankaa. (Preston also gives the fuller compound asafofrankaa).

  27. Trond Engen says

    Hat: Thanks very much for explicating the kenning so thoroughly!

    From me too, and doubly so. I was wondering if I had to try dissecting the kenning. Good thing I didn't.

  28. David Eddyshaw says

    Preston also gives the fuller compound asafofrankaa

    The ɔ is a singular prefix (plural “flags” is afrankaa.) You’d expect it to be dropped in the second element of a compound (as in e.g. Asantehene, the ɔhene “king” of the Asante/Ashanti.)

    Some words do have a zero singular prefix, though, so variants are quite likely on first principles, especially if the word is a loan, which looks very probable. (These prefixes ultimately go back to proto-Volta-Congo noun class affixes, of course, though the associated grammatical agreement system is long dead in Akan.)

  29. Sources that give the zero-prefix singular (in these English-language art / textile collecting contexts) give a plural mfrankaa. (I think ɛ- and ɔ- prefixes imply a- plural, but ∅ is sometimes a- and sometimes m-. But probably even that’s an oversimplification.)

  30. David Eddyshaw says

    I don’t think anyone could fault an intrepid WP editor who just substituted frankaa for the evidently-wrong “frank.”

    There’s been a great deal of falling-together and remodelling of the class-prefix system in Akan. The fairly closely related Guang languages have preserved the system a lot better (at least, Gonja has: it’s the only one I’ve got much of a description of.*) The ɔ/a pairing continues the old “human” gender formally, I think, but it’s got much expanded.

    Gonja still has e.g. ènyɛ́n “man”, bànyɛ́n “men”, recognisable as the same pairing as e.g. Mooré nédà “person”, nédbà “people.”

    * Though even that, Colin Painter’s grammar, is made so opaque by its tagmemic orientation that reading it is like data mining more than reading a grammar. He does slag off Chomsky (politely) in the intro though. Plus points for that, especially in 1970.

  31. Turns out Rattray, in his Ashanti Proverbs, which has come up here from time to time, translating and annotating Christaller’s #2901, did note the probability of English flag as being the source.

  32. #802 is pretty grim.

  33. David Eddyshaw says

    Reminds me of the Kusaal proverb

    Azi’ kpinam kpiidnɛ ka tɛnbid.
    “Those who’ve never died are dying with a struggle.”

    which actually means (something more like) “It’s a lot of fuss about nothing, a storm in a teacup.”

  34. John Cowan says

    Royal Statist. Society

    Is there also a Royal Anarchist Society?

    “A car does not run without petrol.”

    Is that an allusion to bribery?

  35. David Eddyshaw says

    Is that an allusion to bribery?

    “… or saying other potentially unwelcome things, so a layer of ostensible plausible deniability saves everyone’s blushes”

    Is there also a Royal Anarchist Society?

    Alas, efforts to found one ran into insuperable difficulties with finding a suitable patron.

Speak Your Mind