Dmitry Pruss, indefatigable researcher into genetics ancient and modern, linked on Facebook to a PNAS paper by Qian Tang, Edward L. Vargo, Intan Ahmad, et al., “Solving the 250-year-old mystery of the origin and global spread of the German cockroach, Blattella germanica” (tl;dr: “We confirm that B. germanica evolved from the Asian cockroach Blattella asahinai approximately 2,100 ya, probably by adapting to human settlements in India or Myanmar”) and e-mailed me to say “I just wrote a short fb post about the solved mystery of origin of the cockroaches, and realized that there are only so-so lame etymologies of the Russian word. Maybe the LH audience can step in & help?” For your convenience, here’s a webpage gathering the fruits of the etymological dictionaries, primarily Vasmer; there seems to be a general idea that it came from Turkic (either a root tar- < taz- ‘flee’ or “from the derogatory use of Turk. tarkan ‘dignitary’”), but as Dmitry says, that’s pretty lame. Any ideas?


  1. I’d add that Russian has a collection of CaCaCan words mostly of Persian origin.
    Caravan, obviously, but also balagan, sarafan, stakan from dostakan “glass”. Of these sarafan and stakan have been reborrowed to Persian, and the exact source of sarafan is not very obvious. I’m also not sure why chemodan got different vowels in literary Russian despite the availability of CaCaCan sources.

    In this context we also have tarakan and baraban “drum” of which tarakan is the most obscure.
    My point here is that this pattern itself is recognisable either for Russians or for speakers of the source language or both, and words of similar origin that match it imperfectly could be in principle “corrected” to match it (thus my perplexion with -e-o-an).

  2. Good point! And my Persian-English dictionary tells me there’s a ترق taraq ‘cracking (noise),’ so taraqān could be a plural. And cockroaches crack when you step on them.

  3. “chEmOdan” – I just realised that after ch- there necessarily should be a front vowel. So chE- is expected.

    @LH, one of my points is opposite, namely that Russians and possibly Turkic speakers could recognise such “unanalisable -a-a-án words” (consciously or not and with awareness of their source – in the times when such words were borrowed, I mean – or not) and even adapt other words by analogy.
    More or less as we recognise English -ing and use it in ways impossible in English.

    Thus the proposed source for sarafan (a family of words like this one) should have -pâ or -pây, not -fan (it is unclear why -f-). If of course it is the source.
    Thus insertion of -a- in tarkan would not be too unusual.

    The other point is that, yes, most such words are Persian in origin.

  4. Basically all of Turkic has tarakan as the form for cockroach, and with very little variation. This makes it look as though it was borrowed (back) into Turkic from Russian. Fedotov’s Chuvash Etymological Dictionary considers Turkic to be the source of the Russian form, but it’s odd that there’s so little variation from one end of the Turkic world to the other. Turkish, notably, does not use tarakan, but instead uses karaböcek “black bug”.

    Another possibility is that it’s related to Turkic tarak “comb”.

  5. Thanks, and I agree with “This makes it look as though it was borrowed (back) into Turkic from Russian.”

  6. In Kim, Kipling speaks of tarkeean as a good curry and uses the word as a secret code : “ there is no caste where men go to—look for tarkeean,”.

    There is a linguistic discussion at

  7. If Kipling’s “tarkeean” is indeed tarkārī, I can only reconcile them by assuming that the handwritten <ri> was misread as <n>. But would such an error have persisted in the published editions of Kipling’s work for a century? And woudn’t he have transcribed the latter part of the word as <ree>?

    Ed.: The Manuscript of Kim. Tldr, the manuscript is hard to access; Kipling revised his manuscripts in several rounds of typescript before publication.

  8. But would such an error have persisted in the published editions of Kipling’s work for a century?

    Sure — I’ve seen similar examples. If the error is obscure enough, the text has to be very lucky indeed to get a sufficiently omniscient editor/proofreader.

  9. See amendment.

  10. To me, the form in Kipling looks related to Hindustani taṛkā تڑکا तड़का ‘heated oil or ghee in which spices, also onion, garlic, etc. are woken up and browned’ to me (from taṛaknā تڑکنا तड़कना ‘to crack, crackle, snap’). A tarka/tadka is often poured on top of a dish as an aromatic finishing touch. For instance, a दाल तड़का dāl taṛkā is a dish of lentils with a tadka poured on top, a simple and iconic dish. In form, the word in Kipling resembles a feminine plural in -iyā̃. Maybe more in the morning. I have to go to bed now.

  11. a दाल तड़का dāl taṛkā is a dish of lentils with a tadka poured on top, a simple and iconic dish.

    Yum! Now I’m feeling hungry. (Can’t get a decent dāl taṛkā in New Zealand: what they serve is just a pool of lightly-spiced oil floating on top of the dāl. You gotta work a lot of flavour into that oil.)

  12. Dmitry Pruss says

    I thought that Google Ngram might shed some light on the cockroaches of the darkness, but most early mentions in Russian books date to 1830s-1850s, where tarakan mostly stands for the black cockroach (the
    “regular” cockroach) while the “German” / “prussak” is also known, but kind of second in importance. A 1852 zoology book explains that the red cockroach is often found on sea ships.
    But a 1800 zoology (translation of a recent book by Aubin Louis Millin) lists only the black roach.
    A 1820 article in Sibirsky Vestnik mentions a discovery red roaches in Transbaikalia and concludes that they invaded with China trade, probably in the same way as red cockroaches in Denmark, described by Linnaeus.

    All in all, we just conclude that the red cockroach invaded the Russian Empire in the early 1800s, but not a bit more about the word’s origin. And of the earlies mentions, most of it is actually to Тьму-таракань 😀

  13. Nahum Stutchkoff’s Yiddish thesaurus lists “prays, prusak, frantsuyz, rus, tarakan, knayper, shabe, shvab, kakarotsh (Am.)” Are those Prussian/French/Russian/Swabian names varieties of the insect, or insulting euphemisms?

    The available etymologies of cucaracha are also unsatisfactory. Corominas, whom I usually find very reliable, can only offer cuco ‘oruga o larva de cierta mariposa nocturna’, leaving the -racha unexplained, and saying both cuco and cucaracha are of sound symbolic origin. Maybe, and maybe Dutch kakkerlak too (though WAry equates it with the literal ‘shit-licker’), and maybe ultimately tarakan as well, all containing these crunchy consonants.

  14. Y, yes, I too wanted to mention Czech šváb (but cf. German Schabe) and Dutch kakkerlak (comparable to Perso-Indic cracking words mentioned above by LH, david and Xerîb: : taraq and taṛaknā).

    Wiktionary also informs me of
    bubašvába – Serbian/BosnianMacedonian, “Swabian bug”
    rus – Romanian, “Russian”
    cafard – French, especially weird, because it is Arabic كَافِر, kāfir.

  15. Slangy Modern Hebrew uses dzhuk. The official word is תִּיקָן tikán, from תִּיק tik ‘sac’ referring to the egg sac but surely echoing the Russian. מַקָּק makák is more common, and is of a medieval source, from a root indicating decomposition and originally referring to bookworms and such destructive creatures.

  16. I love bubašvába — what a great word!

    …And I just noticed it has the same vowels as cucaracha.

  17. была б баба Люба – она б дала!
    Or rather
    бла б баб Люба – она б дала!
    (the Russian rendering of what is sung in the beginning of Tutti Frutti)

  18. “Serbian/BosnianMacedonian” – for Croatian it gives

    žȍhār – with a strange explanation: “Perhaps from Yiddish שחור (shokher, “black person”), from Hebrew שחור/שָׁחֹר (shakhór, “black”)..”

    Tamil கரப்பான் • (karappāṉ) இரைப்பு (iraippu), கக்கலாத்து (kakkalāttu), இலட்சுமிவண்டு (ilaṭcumivaṇṭu) – kakkalāttu reminds kakkerlak.

    Finnish russakka is from < prussak “Prussian”, not from “Russian”:)

  19. Tashelhit: ibnḍr m – I was not able to find it:-/

    Maltese: wirdien “From Arabic بِنْت وَرْدَان (bint wardān).”
    Iraqi Arabic: صرصر m (ṣurṣur), بنت وردان m (bent wurdān)

    Well, bint is bint and wardaan is…. Biberstein-Kazimirski says “1. Canal. 2. Werdan, n. pr. d’homme et de lieu.“, which makes sense, but again I can’t find the word elsewhere other than in the second sense:/ Or wait… Is it just an untranslatable personal name here? Either one vaguely associated with water because w-r-d or (similarly to “Prussian” etc.) an actual person thought to be the source of cockroaches?

  20. kakkalāttu and kakkerlak – or can the latter – and cucaracha too – be the former?

    The cockroach is said to be “Indian” in a 18th century book (but I don’t find mentions of India in earlier editions of the same book).

  21. Alsoit would be nice to know the chronology of spread of various species.

    Cf. Linnaeus (Systema Naturae. 12th edition, 1767) about the “oriental” cockroach (and if tarakan could refer to any other earlier insect, we are unaware of this): hodie in Russiae adjacentibus regionibus frequens; incepit nuperis temporibus Holmiae 1739 uti dudum in Finlandia. and Sulzer, 1761.: in den Gegenden Russlands, unlängst sind diese ungebettene Gäste auf ihrer Reise nach Westen, nach Finnland, Schweden, Deutschland und Schwaben gekommen. (a collection of early descriptions here)

    Then it could be a recent arrival in Russia as well.

  22. ṣurṣur and the like originally mean ‘cricket’.

  23. @Y: That sounds like it should be related to susurration, but it’s apparently unrelated (except insofar as they both involve reduplication).

  24. I wonder if other onomatopoeic cockroach words originally belonged to entirely different (sound making) species.

  25. Dmitry Pruss says

    Passing the once onomatopoeiic word to a new pest sounds quite likely!

  26. Are those Prussian/French/Russian/Swabian names varieties of the insect, or insulting euphemisms?

    “bubašvába” and “prussak” existing elsewhere make me think that at least the swabian and prussian nods are probably just different regionalisms/etymologies turning up, but those might correspond to different species (along the lines of what Dmitry found). but stutchkof is classically thesaurian in his disinterest in helping disentangle regional lects, varied registers, and specific connotations for those of us who didn’t grow up in the 1930s or earlier.

    refoyl says that “shvab” can be a pejorative for a german as well as a word for cockroach (he doesn’t say whether it could apply to, say, saxons or franconians as well as swabians), but doesn’t otherwise add much to the mix. and, true to form, the schaechter lineage at verterbukh.org only list “tarakan” and “prays”, avoiding the whole question and making a great deal of past writing unintelligible to anyone with only their dictionary to rely on.

  27. While y’all contemplating etymology of tarakan I highly recommend listening to this “romance”.

  28. David Marjanović says

    “Die Deutsche Schabe (Blattella germanica), veraltet auch Schwabenkäfer, Preuße, Russe, Franzose u. a., und wie die Gemeine Küchenschabe (Blatta orientalis) auch als Küchenschabe (Kakerlake) bezeichnet, […]”

  29. Swabian Kafir!
    A double insult to the humble noble animal!

  30. David Marjanović says

    Heh. No, it’s the West Germanic “beetle” word, relictually found in English as chafer (…with the second round of consonant stretching, I guess, like better).

  31. DM, oops.

    Then “Francized form of Arabic كَافِر (kāfir, “unbeliever”).” for cafard must be wrong!

  32. French WAry goes into the history of cafard in much detail.

  33. @Y, thanks! So it is attested in the sense “bigot” slightly before the “cockroach” [and not centuries after, as I thought for some reason].

    Then kafir makes sense semantically. The question is why it would become so popular in French.
    I tried to look for kafir in Spanish.. Coromines and Pascual’s dictionary does not even seem to have cafre, Corriente connects its borrowing with slave trade, (pp 240-1 and 153). If in French it meant un nègre it would make sense for the oriental cockroach is black….

  34. David Marjanović says

    …Looks like “beetle” got mixed up with “drool” somehow, and that’s why the first vowel and the following consonant are such a mess (even just within German, let alone the rest of West Germanic). ~:-|

  35. Somebody should fix that Wiktionary page — under “Middle Low German” we find “German Low German.”

  36. David Marjanović says

    I actually wonder if that’s deliberate – “modern Low Saxon in German spelling”, as opposed to “modern Low Saxon in Dutch spelling” as found in Groningen.

  37. Thanks to Y for pointing out the etymology of French cafard!

    This lends some semantic typological support to Räsänen’s proposal of a Chuvash etymology of Russian таракан. The following is the entirety of Räsänen’s proposal (which Fedatov repeats and which Vasmer also references), available on JSTOR here:

    таракан ‘Schabe’ < tschuw. tar-agan ‘Entflieher, Flüchtling’ ~ türk. täz ‘entfliehen’

    The Chuvash word таракан is perfectly transparent: тар- is ‘flee’, and -акан is the usual and regular present participle ending. Intervocalic -к- is regularly lenited between vowels.

  38. I was wondering if any more semantic typological support could be found to support Räsänen’s etymology as ‘fleeing, fleer’. We might also consider the following entry from an Old English glossary of Latin words, Harley 3376:

    Blatta. lucifuga. wicga. uel genus purpure. uel uermis.

    I would link to a scan of the manuscript page, but British Library site is still down from the October cyberattack. However, this paper has a snippet of a scan of the manuscript on page 581, as well as discussion of philological details of the gloss. See also Wright’s edition here, column 196.

    In any case, this gloss is of interest because of:
    (1) the association between the blatta ‘cockroach; clothes-moth, wax moth; book-worm’ and the tendency to flee (lucifuga)
    (2) the naming of an insect from its motion: OE wicga, probably related to ModE wiggle, and surviving in ModE earwig.
    I suspect this entry in the glossary originated in some connexion to Vergil, Georgic 4.242–247:

    nam saepe favos ignotus adedit
    stellio et lucifugis congesta cubilia blattis
    immunisque sedens aliena ad pabula fucus
    aut asper crabro imparibus se immiscuit armis,
    aut dirum tiniae genus, aut invisa Minervae
    laxos in foribus suspendit aranea casses.

    For often a newt has nibbled
    the combs unseen, wax moths(?), light-fleeing, fill the cells,
    and the useless drone sits down to another’s food:
    or the fierce hornet has attacked with unequal weapons,
    or the dread race of moths, or the spider, hated by Minerva,
    hangs her loose webs in the entrances.

    Compare Isidore Etymologiae12.8 (‘De minutis volatilibus’ (‘Of tiny flying animals’)):

    Blattae a colore nuncupatae, siquidem et conprehensae manum tingunt; unde et blatteum colorem dicunt. Hoc autem animal lucem videre non patitur, contrarium muscae, nam musca lucipeta, et blatta lucifuga est; per noctem enim tantum ambulat.

    Moths (??; blattae) are named from their color, since indeed they stain one’s hand when they are caught, whence people also name a color ‘purple’ (blatteum). This animal cannot bear to see light, in contrast to flies, for the fly is light-seeking and the moth is light-fleeing, and it only goes about at night.

    (I would have intially thought that the blatta intended was ‘cockroach’ rather than ‘moth’ here, because of the disgusting ‘purple’ left behind, but the next animal discussed is papilio ‘butterfly’, and cockroaches in general aren’t great fliers. Draw your own conclusions.)

    In any case, these texts show how an epithet like lucifugus might be set up to displace another word as a name for an insect, similar to what Räsänen proposes.

  39. In regard to the etymology from Turkic tarqan, I am waiting to get my hands on the work Vasmer refers to, Roman Jakobson (1959) ‘Marginalia to Vasmer’s Russian etymological dictionary’, International Journal of Slavic Linguistics and Poetics 1 (2), but I am travelling for the Eid holiday. In particular, I am wondering why cockroaches would have been associated with a Central Asian tarkhan. (The cockroaches’ усы or усики ‘mustachios, whiskers; antennae, feelers’, likened to some elaborate drooping mustachios thought of as being typically cultivated by Central Asian worthies like tarkhans? Characterization of tarkhans as fleeing cowards?)

  40. “modern Low Saxon in German spelling”, as opposed to “modern Low Saxon in Dutch spelling”

    > clutches head <

    And which of those is “German Low German.”?

    As if anyone would confuse "East Pomeranian Dialects" with "Mecklenburg West Pomeranian Dialects" with "West Prussian Dialects [MORIBUND]". Henry Higgins would tell a Dustman from Stralsund vs a Chimneysweep from Güstrow.

  41. David Marjanović says

    > clutches head <

    They ended up having separate Wikipedias, you know.

    And which of these

    The one in German spelling of course.

  42. ktschwarz says

    David M deduced correctly: German Low German is the en.wiktionary name for Low German spoken in Germany, as distinguished from Dutch Low Saxon. (Both pages are labeled “draft proposal” and “unofficial”, though.)

  43. OK then! It’s just random weirdness that I’ll have to get used to. But why aren’t “Low German” and “Low Saxon” sufficiently distinct on their own?

  44. On a tree by the Weser a little tom-tit sang “German Low German, Low German”

  45. David Marjanović says

    But why aren’t “Low German” and “Low Saxon” sufficiently distinct on their own?

    Because the versions spoken in Niedersachsen are very much Low Saxon, too…

  46. Someone must name something “Nederlands Nedernederlands”.

  47. Someone named Lekkerkerker, perhaps.

  48. We have “the counsil of the federation of the federal assembly of the Russian federation” (usually translated as “the Federation Counsil”).
    I did not even know this version of the name until I saw it in large letters on the building.

Speak Your Mind