Recent Focused Series »

Indo-European Origins
Siberia
Northern California
The Caucasus
Imaginary Geography
Home » Cultural Geography, Europe, Indo-European Origins, Linguistic Geography

The Geography of the “Onion” Vocabulary

Submitted by on January 30, 2013 – 6:06 pm 41 Comments |  

[Thanks to Diane Nelson, David Pesetsky, Steven Lubman, Alex Dratva, Alexei Kassian, Hsin-Chang Chen, and Ruta Nylander for their help in researching this post.]

onion_map

As pointed out in an earlier GeoCurrents post, examining the history and geography of just one word across languages can reveal fascinating and instructive patterns. In this post, we will take a closer look at the words for ‘onion’—as well as its relatives, leek, garlic, scallion, and shallot—in a number of European languages. Starting with ‘onion’, the GeoCurrents map on the left reveals three main groups of cognates, but crucially, none of them correlates with main subgroupings within the Indo-European language family (or with the family as a whole), even though ‘onion’ can be thought of as a pretty basic word. Closely related languages may have very different words for ‘onion’, while distantly related or even unrelated tongues may feature cognates for this meaning. The important lesson to draw from this is that the distribution of cognates for any single meaning (and by extension a relatively small set of such meanings, such as a Swadesh list) may tell an interesting story, but it is often one of both common descent and borrowing.

The largest group of cognates on this map (shown in blue) are words related the Latin cepa ‘onion’, of unknown origin. The closest descendants of the Latin cepa are the Catalan ceba and Romanian ceapă, while Italian cipolla, Portuguese cebola, Galician sebola and Spanish cebolla derive from Late Latin diminutive cepulla (literally, ‘little onion’). While many Romance languages, which descend from Latin, retained the original root, not all members of the family did. For example, the French word is oignon, to which we return below. On the other hand, not all the cognates of cepa/cipolla are found in languages of the Romance grouping. This root caught on in German, a Germanic language, though it was altered by folk etymology in Old High German (zwibolla) as deriving from words for ‘two’ and ‘ball’, with the resulting form in Modern High German being Zwiebel. The cepa/cipolla root has also penetrated West Slavic languages—Polish, Czech and Slovak—which have cebula, cibule and cibul’a, respectively. East Slavic Ukrainian, which is known for its active and bidirectional lexical exchange with Polish, has tsibulja. One of the South Slavic languages, Slovenian, which is located close to Italian, also has a cipolla-like word, čebula. Latvian, one of two Baltic languages (which are usually analyzed as most closely related to the Slavic grouping), also features a cepa-cognate, sīpols, while its closest relative, Lithuanian, had a similar word until early 20th century (we will return to Lithuanian below). North of Latvian, Estonian and Finnish—both members of the Finno-Ugric family—too have borrowed the same root (their respective words are sibul and sipuli). (Note that another Finno-Ugric language in Europe, Hungarian, uses a completely unrelated word, hagyma, for ‘onion.) Finally, Basque—also a non-Indo-European language— borrowed its ‘onion’ word tipula from Spanish.

As mentioned above, the French did not keep the cepa root, at least not for ‘onion’ in general. Instead, they opted for another Latin root, “repurposing” unionem (nominative form unio) meaning ‘union’. This word was used in its ‘onion’ meaning already in colloquial Vulgar Latin, presumably because layers of an onion keep together as a bulb. In the early 1100s, Normans brought this word across the channel, giving rise to the English onion. However, the use of unionem-related word for a member of the allium family was not novel for the English since Old English had ynne (as in ynne-leac), from the same Latin source. Elsewhere in the British Isles (see cognates in pink on the map), this Latin root produced Irish inniun (and later oinniún) and Welsh wynwyn (and winwnsyn). Back on the continent, in Dutch, the ending in -n was mistaken for a plural inflection and new singular ui was formed. Thus, the ‘union’-based word for ‘onion’ took hold in some Celtic and Germanic tongues, in addition to French. However, the French did not abandon the Latin cepa-root entirely, deriving from it the word ciboule, referring to ‘green onion’, ‘spring onion’ or ‘scallion’. The trace of the Latin cepa is also retained in the English word chive, borrowed from French in the 13th century.

The third group of ‘onion’ words (shown in brown on the map), containing variation on luk, is found in North Germanic languages (Danish løg, Swedish lök, Norwegian løk, and Icelandic laukur), as well as in two of the East Slavic languages (Russian and Belarusian) and some South Slavic languages: Serbian, Croatian, and Bulgarian. In all of these languages, ‘onion’ is luk. The non-correlation of familial groupings and cognates for ‘onion’ is obvious here, with one of East Slavic languages (Ukrainian) and two of the South Slavic ones (Slovenian and Macedonian) being dissenters. Although Macedonian is most closely related to Bulgarian (so much so that many Bulgarian scholars deny it the status of a separate language altogether), its ‘onion’ word, kromid, is similar to the Greek kromion. The luk cognate set is said to derive from Proto-Germanic form *lauka- though it is not clear whether it referred to ‘onion’, or alternatively to ‘leek’ or even ‘garlic’. According to a renown etymologist Max Vasmer, who favored the Germanic superstratum theory for Slavic languages, the luk words in East Slavic languages derive from the linguistic influence of Varangians (Vikings), who opened trade routes in Russian lands in the 9th century and later established the ruling Rurik Dynasty. Since the same root is thought to have penetrated South Slavic languages via Old Church Slavonic, it is also possible that Russian (and thence Belarusian) got this word from Old Church Slavonic, alongside numerous other borrowings, including both learned and everyday words. Going back to Germanic, Old English had a related word, recorded as læc in Mercian dialect and leac in West Saxon, and it too was an umbrella term for ‘leek’, ‘onion’, and ‘garlic’; the Modern English word leek and the Modern German Lauch (also meaning ‘leek’) derive from that root, as does Dutch look meaning ‘leek’ or ‘garlic’. Speaking of garlic, this English word grew out of the same root. It is recorded for Old English as garleac (in Mercian dialect) and garlec (in West Saxon dialect), and is a compound of gar ‘spear’ and leac ‘leek’.

Another fascinating twist to the ‘onion’ story comes from Lithuanian. In earlier times, the cepa-related cibulis was used, but around 100 years ago it was replaced by svogūnas, a form that is unlike anything else seen in Balto-Slavic languages, or indeed any Indo-European languages in Europe. It appears to have been borrowed from Turkic languages: compare with the form sogan (with “g” in all Turkic forms pronounced as a velar fricative), found in Turkish, Azeri, Turkmen, Chuvash, Nogai, and Kumyk, as well as the Tatar sugan, Kirghiz sogon, Bashkir hugan, Balkar soxan and a score of other similar forms in other Turkic languages. However, it does not appear likely that Lithuanian borrowed directly from a Turkic tongue; rather it is probable that Finno-Ugric languages spoken in the heartland of Russia served as an intermediary. As pointed out above, Finnish, Estonian, and Hungarian all have unrelated words, but sogan-like forms are found in Komi-Permyak (sugon’ ‘onion’) and other Finno-Ugric languages of the region.

To round up our exploration of the ‘onion’ vocabulary, consider English words for two other related culinary varieties: scallion and shallot. The former makes its appearance on the British shores around 1300, deriving from Old North French escalogne (alternatively, from Old French eschaloigne), which in turn grew out of the Vulgar Latin *escalonia, an abbreviated form of (cæpa) Ascalonia, literally ‘(onion) from Ascalon’, an ancient Philistine seaport described by the Greek writer Theophrastus and known today as the coastal Israeli city of Ashkelon. The word shallot came into English also from French, but in a later period, probably the mid-1600s, as witnessed by the appearance of the sh-sound. By that time, the French form was eschalotte, but the origin of this word is in the same Vulgar Latin *escalonia.

Previous Post
«
Next Post
»

Subscribe For Updates

It would be a pleasure to have you back on GeoCurrents in the future. You can sign up for email updates or follow our RSS Feed, Facebook, or Twitter for notifications of each new post:
        

Commenting Guidelines: GeoCurrents is a forum for the respectful exchange of ideas, and loaded political commentary can detract from that. We ask that you as a reader keep this in mind when sharing your thoughts in the comments below.

  • RIJO P GEORGE

    In Kerala, India onion is called “Savola” while in the neighbouring Tamil Nadu it is “Vengayam”. Now I know that “Savola” came from the Portuguese

    • http://www.pereltsvaig.com Asya Pereltsvaig

      Thank you for this fascinating addition! I didn’t even venture into the non-European branches of IE.

  • Pingback: Nibbles: Yarsagumba, Chocolate meet & dig, Beer dig, Mapping disease, Mapping language, Going digital, Urban ag meet, Weird citrus, CGIAR genebanks

  • grandepuffo

    The Turkish word for onion is “sogan” (g is not pronounced). There is one word on the map: “svogunas” which looks similar. I can’t tell which language is the latter, but perhaps “sogan” also somehow derives from the Latin “cepa.”

    • http://www.pereltsvaig.com Asya Pereltsvaig

      Yes, I mentioned the Turkish word in the post, and it doesn’t seem to be derived from Latin…

  • Steve

    Two observarions. First, Basque “tipula” comes from earlier “kipula”, which is thought to derive directly from Latin – at a time when it still had a “hard” c [k]. So it’s certainly not a Spanish borrowing (and if it were, it would probably have an initial z or tz).

    Also, I’m not sure it’s that unlikely that Lithuanian would have borrowed their onion word directly from Turkic. Until some time ago there was a relatively large Lithuanian Tatar (Lipka) population there (Charles Bronson was one famous descendant of Lithuanian Tatar migrants). Couldn’t the “new” Lithuanian onion word come from them?

    • http://www.pereltsvaig.com Asya Pereltsvaig

      Good point on the Basque, Steve! As for exact Turkic source of the Lithuanian word, several different groups have been mentioned, but we don’t know for sure, and none seems to be an appropriate candidate from a sociolinguistic point of view. So it’s still rather a mystery, I’m afraid…

  • LEX

    Here’s a film about the onion’s etymology in English.

    http://www.youtube.com/watch?feature=player_embedded&v=0mbuwZK0lr8

    • http://www.pereltsvaig.com Asya Pereltsvaig

      Thanks, LEX! Nicely animated but completely wrong when it comes to tge historical spelling of ONION, I’m afraid. Pity, I liked how they animate morphological concepts and wish I could use it in class…

      • LEX

        Actually, the film isn’t wrong; I’m afraid you’ve misapprehended its argument. The film doesn’t make any claims about the historical spelling of ONION. It makes a claim about the synchronic, morphological analysis of the spelling of ONION. The word ONE (with an Old English root) and the word ONION (with a Latinate root) are, of course, cognate. But the film doesn’t claim that the word ONION is etymologically derived from the word ONE; it claims that the word ONION is, in the present day, morphologically derived from the BASE ELEMENT spelled O-N-E.

        • http://www.pereltsvaig.com Asya Pereltsvaig

          Perhaps I misunderstood the animation, but what does it mean “morphologically derived from the BASE ELEMENT spelled O-N-E”? Historically, ONION and ONE do derive from a common source albeit very indirectly, and most certainly ONION does not derive from ONE (which I indeed understood the animation to claim). Synchronically, the word ONION does not contain morpheme ONE either, anymore than ELEVEN does. So while the animation is trying to make a valid point, it is not phrased in a linguistically sensible way, I am afraid.

          • LEX

            In orthographic linguistics, an element is a written morpheme, identified only by its spelling and meaning and not its pronunciation. A base element is the main written morpheme of a word. An element is identified by its spelling because it has no pronunciation until it surfaces in a word. The base element may be free or bound. The written word ONION does contain the base element O-N-E, and the final E drops when the vowel suffix is added. The spelling is coherently analyzed as ONE + ION, but these are spellings — I can’t use angle brackets in this comment, but would to signal the orthography. The written word UNION is the base element UNE (seen also in TRIUNE) + -ION. TENSION is TENSE + ION. It’s a question of the written structure of the word, not a question of what native speakers are or aren’t aware of. The word ELEVEN is, of course, historically related to the word ONE, but they don’t share a spelled element; ONE and ONION do. While many linguists focus only on spoken language, orthographic linguistics considers the writing system as a coherent system itself. The film is about spelling, and is “linguistically sensible” in its analysis of the orthography. It makes no claims about the morphology of the spoken language. A notion of morpheme that is constrained to spoken language will not be sufficient to cultivate an understanding of the orthographic structures. Morphological derivation, of course, involves deriving a new lexeme from a morphological process (like, in this case, affixation).

          • http://www.pereltsvaig.com Asya Pereltsvaig

            Orthography represents spoken language, not the other way around, and the two are not independent systems. So why some aspects of orthography make sense linguistically, others are a pure convention and therefore do not represent anything “linguistically sensible”. For example, spelling of “friend”—why the letter I?

          • LEX

            Why the I in FRIEND? It’s an etymological marker letter (like the W in TWO and the B in DOUBT) — it marks a connection to its historical root, and to the word ‘Friday,’ which shares both a historical connection and the base element FRI, denoting ‘love.” Indeed, the French ‘vendredi’ and Spanish ‘viernes’ and Italian ‘venerdi’ are named for Venus; English ‘Friday’ and German ‘Freitag’ and Dutch ‘vrijdag’ are all named for Frigga (or Freya), the Norse goddess of love. A FRI + END is one who is loved, just as a REVEREND is one who is revered.

            Have you ever heard of a Man Friday or a Girl Friday? This is a trusted helper, a (business) friend. It’s no coincidence that Robinson Crusoe met his friend on Friday, nor that Friday became that friend’s name. Such etymological connections were not lost on Defoe or many of his readers.

            That’s why the I in FRIEND. It makes a lot of sense, as long as we don’t assume that phonology is the only (or even the primary) consideration in how the writing system works.

            Of course written and spoken language are not independent; this was not my claim. English was born with a writing system already in place, and in the present day, our orthographic knowledge affects our speech and our perception of it from our childhood on. Orthography’s purpose is to represent sense and meaning. Written forms may ‘represent’ (morpho)phonological forms, but that doesn’t mean that orthography represents pronunciation. It’s more accurate to say that written forms (elements) may be realized in speech (morphs).

            English orthography makes sense — that’s its job — and it does so linguistically on several levels: morphologically, etymologically (both synchronically and diachronically), and phonologically.

          • http://www.pereltsvaig.com Asya Pereltsvaig

            I’m afraid you are wrong on the spelling of FRIEND: historically there was no letter I in it. Adding it, or the S in island, is a case if folk etymology.
            http://www.etymonline.com/index.php?term=friend&allowed_in_frame=0

          • LEX

            No, I’m not wrong about the spelling of FRIEND. You missed the evidence. The Online Etymology Dictionary is an excellent resource, and one of the most excellent things about it is its cross-references. The link you posted, for FRIEND, says “see FREE” — and when you link to FREE, you will see information that verifies my assertions, but you have to actually read the whole thing.

            Moreover, etymonline is not the ONLY resource, and every good scholar interrogates more than one source, of course. Doing so would likely help you discover that Old English spelling did not have a single, fixed form. Nor did Middle English spelling. English spelling did not begin to become conventionalized and fixed until the Modern period. As it evolved, the system opted for the form of FRIEND with an I because it was the optimal form for representing meaning and sense to those who know the language. Note how many of the historical forms of FRIEND attested in the Oxford English Dictionary indeed do have the letter I in them:

            OE fréond, fríond, fríend, frýnd, ME friend, vriend, frend(e), vrend(e), freond, vreond, freind(e), vrind, vryend, freend(e, freynd, frind(e), frynd(e), freyind), 15th friend, frend, friend(e), frond, freond.

            Etymonline also gives us the close-to-Old-English forms of Old Frisian FRI and Old Saxon VRI.

            The S in ISLAND was indeed a case of folk etymology, as it is not at all related to Latin insula, as people mistakenly thought. There are a handful of words whose etymology indeed includes folk etymologizing; that’s part of the story of those words. But it’s not part of the story of FRIEND.

            I’m offering an etymology seminar this spring with the author of the Online Etymology Dictionary. There’s information here if you’d like to come: http://linguisteducatorexchange.wordpress.com/2012/10/23/a-weekend-to-remember/

  • Pawel

    Lithuanian “svogūnas” actually comes from Karaim “sogan”. The Karaims – settled down since the end of the 14th century in Trakai – were noted gardeners that developed local horticulture to the high standards.

    • http://www.pereltsvaig.com Asya Pereltsvaig

      Thanks for the comment, Pawel! I’m not sure about the Turkic source for the Lithuanian word though, as I pointed out in another comment below, for sociolinguistic reasons. There are several theories but none seem completely proven to me…

      • n

        Both of you are right, Karaim is a Turkic language :)

        http://en.wikipedia.org/wiki/Karaim_language

        • http://www.pereltsvaig.com Asya Pereltsvaig

          Oh I know that: http://languagesoftheworld.info/russia-ukraine-and-the-caucasus/crimean-karaim-who-are-they.html

          What I don’t agree with is the idea that Karaim are the group who contributed the loanword for ‘onion’ to Lithuanian: they were too insignificant and unprestigious from the Lithuanians’ point of view…

          • Pawel

            You are right: they were insignificant and unprestigious but not for the late 19th and the early 20th century Lithuanian lexicographers who created modern Lithuanian language and deliberately chose this word to replace – too Slavic for them – “cibule” and “cibulis”.

          • Pawel

            Writing “too Slavic” I meant “too Polish”, of course.

  • http://www.facebook.com/profile.php?id=1169599108 Mike Campbell

    From my experience teaching etymologies and combining geography in such a way, which I love, I would suggest abandoning each language’s local orthography and re-writing the words on the map in broad transcription IPA so that comparisons are more easily identified by those who can’t read each language. In order to make your case (for students who don’t fully understand diachronic morphophonology) you could re-write the IPA in broader terms, such as features, so that affricates, palatals, etc line up with the correspondences in other languages.

    • http://www.pereltsvaig.com Asya Pereltsvaig

      Thanks, Mike! I was wondering whether to do the map in the original orthography, traditional transliteration or IPA. But since many of our readers may not be familiar with IPA, I decided against it. So the map is a mixture of traditional orthographies and traditional transliteration (where appropriate).

  • Matthieu

    In South of France, in Occitan language “the Onion” is “la Cèba”. There is a village near “Montpellier” called “Lezignan la cèbe”. This village traditionally grown onion.

    • http://www.pereltsvaig.com Asya Pereltsvaig

      Fascinating, thanks for sharing this info, Matthieu! As several readers point out, more can be added to the map, although the overall point remains the same.

  • billposer

    In Japanese “scallion” is negi and “onion” is tamanegi “ball scallion”. BTW, with regard to Turkish soğan, although as a previous commenter mentioned the ğ is not pronounced in Standard Turkish, historically it was indeed a voiced velar fricative as it still is in many Turkish dialects.

    • http://www.pereltsvaig.com Asya Pereltsvaig

      Very interesting information about Japanese, billposer. And thanks for the clarification about Turkish pronunciation (I’m sorry about my bad wording in the original post).

  • http://irinarempt.pip.verisignlabs.com/ Irina

    Frisian has ‘siepel’ and several northern/eastern Dutch dialects have similar forms; perhaps from rubbing against Germany.

    • http://www.pereltsvaig.com Asya Pereltsvaig

      Thanks, Irina! This is fascinating!

  • Christopher Culver

    While Hungarian hagyma has no cognates in Europe, this lexeme is also found in Mansi as kośśəm.

    • http://www.pereltsvaig.com Asya Pereltsvaig

      Thank you, Christopher! An interesting tidbit, but to be expected, as Mansi (and Khanty) are linguistically the closest relatives of Hungarian…

    • TʀoᴘʏʟıuM

      I believe Udmurt /kumɨz/ and Komi /komidʑ/ (both of which I think refer to garlic) have also been considered cognate, despite an apparent consonant metathesis. An approximate Proto-Uralic shape *kanśəma or *kaćəma could be set up — though perhaps there are better explanations (after all, this post demonstrates that names for “onion” are prime loan material).

      Another European subfamily of Uralic seems to display a yet different root for the plant: Erzya /tʃurʲka/, Mokša /ɕurʲkɛ/. I have no idea on the further etymological background of this, but it appears unrelated to all words brought up so far.

      • http://www.pereltsvaig.com Asya Pereltsvaig

        Fascinating, thanks!

  • Diell

    Hey Asya,

    You mentioned the closest descendants to the Latin “cepa” being Catalan, and Romanian.

    The Albanian word for onion, is also directly derived from the same source in the form of the word “qepë”. The “q” represents a voiceless palatal stop.

    Thanks for the interesting post!

    • http://www.pereltsvaig.com Asya Pereltsvaig

      Thanks, Diell! that’s fascinating! Seems like the Latin word for ‘onion’ really got around, didn’t it?

  • Toni Keskitalo

    ‘Onion’ in many languages can be found on Gernot Katzer’s spice pages: http://gernot-katzers-spice-pages.com/engl/Alli_cep.html

    If I’m not mistaken, North Saami “lávki” is from Scandinavian but Skolt Saami “luukk” is from Russian (which is again from Germanic).

    Also, thank you for wonderful articles here!

    • http://www.pereltsvaig.com Asya Pereltsvaig

      Thanks for the link to the spice pages and the info on Saami—fascinating!

  • s/o

    So if all the slavic languages either borrowed the late latin word or took the germanic root via the vikings and old church slavonic, is there no native slavic word for the vegetable? Or are there other names alongside or replaced by the borrowed ones? (Or did they just not have onions?)

    • http://www.pereltsvaig.com Asya Pereltsvaig

      Good question! As far as I can tell, they did have onions. But it appears that the borrowing of LUK into Slavic happened earlier than I said in the post, quite probably from Proto-Germanic into Proto-Slavic (with some Slavic languages later replacing it with another form). If so, perhaps Slavic languages never had a distinct form for ‘onion’…

  • Pingback: News Roundup: Education, Equality, Health, & Human Rights « Wherever I am, you are there also

  • Alex Fink

    There’s a nice reconstruction due to Witczak [http://www.akademiai.com/content/k84583228lx77600/] suggesting that in fact the Latin _ūniō_ continues a PIE term *wósHr/n for some member of genus Allium, and fell together with _ūniō_ ‘oneness’ which it was then folk-etymologised as being an extended use of. This *wósHr/n would be found also in Hittite _šuppi-wašḫar_ ‘onion’ (first element meaning ‘pure, holy’), Sanskrit _uṣṇa_ used for ‘onion’ but with primary meaning ‘hot’ (though rather ‘sweltering’ than ‘pungent’), and some Indo-Iranian ‘garlic’ (Pashto _ūža_, Khowar _wǝẓnū_).

    I came across this at http://polyglotveg.blogspot.com/2007/03/garlic.html , part of a great series of similar nature to this one, though sadly sans maps.

    • http://www.pereltsvaig.com Asya Pereltsvaig

      Thank you for your comment and the link, Alex. I am not too convinced by this alternative etymology though… And yes, the Polyglot Vegetarian series is very interesting!