Loan Words.

I’m so used to news media having uninformed pieces on language that it’s a pleasure to find exceptions; BBC News had a magazine story on loan words by Philip Durkin, who — being deputy chief editor of the OED — is definitely up to the task, with interesting tidbits like this:

Today English borrows words from other languages with a truly global reach. Some examples that the Oxford English Dictionary suggests entered English during the past 30 years include tarka dal, a creamy Indian lentil dish (1984, from Hindi), quinzhee, a type of snow shelter (1984, from Slave or another language of the Pacific Coast of North America), popiah, a type of Singaporean or Malaysian spring roll (1986, from Malay), izakaya, a type of Japanese bar serving food (1987), affogato, an Italian dessert made of ice cream and coffee (1992).

I found the odd-looking quinzhee in the OED and discovered it’s pronounced /ˈkwɪnzi/ (QUIN-zee, just like the traditional pronunciation of Quincy) and is from “Slave kǫ́ézhii, lit. ‘in the shelter’, or < a similar form in another Athabaskan language." And they ran a followup piece in which Durkin “looks at readers’ own favourite examples”; there are lots more goodies there. A sample paragraph:

It is often useful to distinguish between immediate and remoter origins of words. For instance, among the French borrowings into English in the original article, peace comes from an earlier form of French paix which goes right back to the Latin origins of the French language (the Romans spoke about pax), but war comes from a northern variant of French guerre, a word which French originally borrowed from a Germanic relative of German and Dutch. A similar example noted by a reader is boulevard, a word that English borrowed from French in the 1760s, but that French itself borrowed in the Middle Ages from Dutch bollwerk or a related word, making the word seem more familiar by substituting the ending -ard of words like placard. In some cases English acts as the middle-man — cake probably came into English from early Scandinavian in the 1200s, but has since been borrowed from English into numerous languages in Europe and beyond.

Unfortunately, the last paragraph goes a bit astray, beginning:

Just sometimes, though, many languages across the world will have similar words in the same meaning for reasons other than borrowing. The clearest example is probably from words for “mother” and “father” around the world that are superficially similar to mama, dada, or papa. Such words all ultimately go back to the sounds that babies throughout the world produce when they first start to master the art of producing distinct speech sounds, the familiar “mamamamama” or “dadadadadada” that few parents can help interpreting as their own special greeting, and that have given rise to many and various words for mother and father all around the world.

Obviously, a far more common reason for languages having “similar words in the same meaning for reasons other than borrowing” is that they are cognate, like French cinq and Spanish cinco. I presume the problem lies in the editing process; it’s just a miracle there aren’t more blunders on that account. At any rate, I encourage BBC News and everyone else to let professionals write their language pieces instead of inserting ignorant reporters into the mix. (Thanks, Paul!)

Update (August 2015): I should have added this infochart with accompanying essay by Durkin back in 2014, but better late than never.


  1. Slave is not or a “language of the Pacific Coast of North America”. That’s Durkin, sloppier than the OED.
    The OED is still lazy when it comes to extra-European etymologies. What’s this “or < a similar form in another Athabaskan language"? Couldn't they be more specific? May I presume this word is not from Navajo?
    Also, G-Books has quinzhee from 1975 and quinzee from 1983 and a bunch of others predating the OED examples, but maybe those were not online yet when the entry was compiled.

  2. Re: Papa/Dada/Baba and Mama and its variants, I have often wondered about the Chinese words 爹/diē and 娘/niáng vs. 爸/bà and 妈/mā. My impression is that diē and niáng are the older, “more original” words for father and mother, while bà and mā look like missionary school loan words, and that impression is reinforced by a friend from Shandong telling me that when he was young, immediately post-GPCR, kids who called their parents bà and mā were accused of “装洋”, or “acting Western”. And diē and niáng seem to defy the “first sounds babies make” theory. Unfortunately, every previous attempt I’ve made at finding out the origins of these words has been confounded by the common insistence that the origin and development of characters is “etymology”, while I insist that etymology is tracing the origin and development of words. And to rub salt in the wound, there are those who insist the term “character etymology” is adequately clear and distinct a term. And I just discovered that my “A Graphic Compendium of Chinese Characters”, which is quite good for this sort of thing, gives me no clue as to the relative ages of these words for father and mother.

    “war comes from a northern variant of French guerre, a word which French originally borrowed from a Germanic relative of German and Dutch.”

    Related to German Heer and Wehr?

  3. Further to the baba/dada/mama thing:
    In Māori, hākui means mother or mum, and māmā is a loan word. That dictionary didn’t mark pāpā as a loan word for dad, though.

  4. John Cowan says:

    diē and niáng

    1) Not all words for ‘mother’ and ‘father’ are mama-papa words, at least not detectably so. Gothic aiþei ‘mother’ is of uncertain origin: it could be < aiþs ‘oath’, or a feminine version of atta ‘father’. In any case, it or some closely related East Germanic word was borrowed into Finnish as äiti (but cf. emo ‘dam’, Estonian ema ‘mother’, of Proto-Uralic origin).

    2) When mama-papa words become part of the adult vocabulary, they undergo sound changes like any other words, which can transform them into something barely recognizable (e.g. English mother, father) or completely unrecognizable.

    Related to German Heer and Wehr?

    The OED says: “Old High German werra (Middle High German werre ) confusion, discord, strife, related to the Old High German, Old Saxon werran strong verb, to bring into confusion or discord (whence modern German wirren weak verb to confuse, perplex; the earlier verb survives in verworren participial adjective, confused), < Germanic root *werz- , *wers- , whence also worse adj.”


    It is a curious fact that no Germanic nation in early historic times had in living use any word properly meaning ‘war’, though several words with that meaning survived in poetry, in proverbial phrases, and in compound personal names. The Romanic-speaking peoples, who were obliged to avoid the Latin bellum on account of its formal coincidence with bello- ‘beautiful’, found no nearer equivalent in Germanic than werra [thus guerre/guerra]. In Old English the usual translation of bellum was gewin, ‘struggle, strife’. The continental Germanic languages later developed separate words for ‘war’: German krieg (whence Swedish, Danish krig), Dutch oorlog; Icelandic uses ófriðr ‘un-peace’.

    From what I can make out, Wehr ‘defense > weir, barrier’ is a deverbal noun from wehren, which may be a -j causative of the same root. Heer has a tentative PIE etymology *koro-s ‘struggle’: Grimm compares it to Skt kula ‘herde, schwarm, menge, geschlecht, genossenschaft’, but the OED is doubtful.

  5. Grimm compares it to Skt kula ‘herde, schwarm, menge, geschlecht, genossenschaft’, but the OED is doubtful.

    The Grimm lemma on Heer is equally doubtful. It merely quotes the doubt expressed in the BÖHTL.-ROTH source (from which the comparison itself comes, not from Grimm):

    nur zweifelnd ist es mit sanskr. kula herde, schwarm, menge, geschlecht, genossenschaft (BÖHTL.-ROTH 2, 351) zusammengestellt worden.

  6. “..ignorant reporters”. To defend my trade – unless they are on a specific beat, reporters get thrown day after day into stories about which they know nothing. News organisations in my time could not afford to have specialists for more than a few select areas – and even less so today. I always said that that a reporter “is an expert in three days or 10 years.” I was the three day version, on subjects ranging from AIDS (try covering one of the first World AIDS Conferences with the most sketchy background) to major oil spills.

    So LH, a bit of pity for the poor hack :-)

  7. Oh, yes, I definitely understand the reporters’ situation, and of course ignorance is no vice — we are all of us terminally ignorant. But it’s hard not to get annoyed when you know about a particular subject and can see the reporter mangling it before your eyes, especially when you know the interviewee was telling the reporter the right thing and the reporter is passing along the wrong thing. Not knowing is one thing; not knowing and not allowing your ignorance to be even momentarily improved by an expert is another. I know, I know, it’s hard and they only have so much time to devote to any given expert and so on, and if I were a reporter I’d have much more sympathy and understanding for their situations, but it’s like drivers and pedestrians: when you’re one you get very irritated with the other, even though you’ve been in their shoes yourself. Anyway, no offense intended to your noble profession!

  8. Then there’s Japanese, in which famously ‘daddy’ is chichi (*titi) and ‘mommy’ is haha (*papa).

  9. marie-lucie says:

    “Mama/Papa” may be the best known baby words, and syllable duplication is quite common, but these formations are far from universal in the languages of the world.

  10. In Iceland the ma or mam sound infants commonly make is generally understood to mean mamma. A rough draft of an important word.

    But the da sound is always interpreted as “datt” (fell or fell down).

    A very small child (imagine one sitting in a highchair) will often hurl things to the floor with much glee — small toys you place before them, cups, plates, food, whatever. This is often accompanied by the da! sound. So parents (when they pick stuff off the floor) will usually respond by saying datt? It fell down? Did it fall down? In a baby-talk way.

    The pa sound gets no attention in Iceland. Perhaps infants quickly learn that the da sound will almost always elicit a response so they don’t bother with the pa sound.

  11. Stu: Böhtlingk & Roth is a Sanskrit dictionary, so they’re only being cited for the meaning of kula, not for the opinion that it’s related to Heer.

  12. John:
    I wasn’t sure about “BÖHTL.-ROTH” being the source of the comparision, Trouble is, that parenthetical expression is the only explicit subject candidate in sentence and context.

    Perhaps the dictionary compilers did not make the comparison (“BÖHTL.-ROTH” does not appear in the source glossary). Then the Grimm sentence is one of those German impersonal-passive statements that make perfect sense, but are rather hard to tease into good English without tripling the number of words.

    It means: “someone has suggested a relationship between x and y while expressing doubts about it”. What the last bit does – “while expressing doubts about it” – is done in German by the single adverb zweifelnd placed where it is.

    In any case, whoever wrote the lemma is passing on, without any commitment on their part, the information that someone or other is dubious about any relationship between kula and Heer. That was my point – that not only the OED is doubtful.

  13. Of course one might say: “has doubtingly considered a relationship between …”, or even “has considered that a relationship between … is doubtful” – but this last version transfers the doubt from the considering to the relationship. Maybe only a nit-picking ontologist would scruple at this.

    When translating into felicitous English I find myself often tempted to grab for metonymy-or-whatever-it’s-called. A critic might scream bloody murder – “that’s not the exact meaning of the sentence !” – but that’s what critics are for.

    In the original sentence the meaning of zusammenstellen is more or less “compare”, but it’s an unusual use. zusammenstellen usually just means “compile” (“systematize”, not IT “compile” !), or just “set down next to each other”. I introduced the words “comparison”, “consider”, “relationship”, “suggest” in an attempt to cover all bases.

  14. @Stu: Maybe “hesitantly”?

  15. Ran: that’s a good nuance to consider. However, “hesitantly” would not rule out a certain expectation that there might be a connection after all. It would characterize the way in which the mooted connection was put forward by the speaker.

    ist zweifelnd … zusammengestellt worden” says that a connection was mooted and characterizes the person who mooted as having doubts it. No hesitation is implied.

    That particular sentence is a bit 19C, I wouldn’t expect anyone to write like that today. It does nevertheless exemplify a standard feature of German – more in its literary than colloquial forms – as regards “adverbs”.

    In this particular example, zweifelnd appears to be an adverb modifying zusammengestellt worden, and would be identified as such by a normal grammatical analysis. The sense of the construction, though, is adjectival – it ascribes doubt to the subject.

    I remember the higher-consciousness-of-German phase I entered when I noticed this and certain other things about German. From then on, I would say things like “Germans do not speak logically, but impressionistically”.

    But a extremely simple analogy in English has just this instant occurred to me – the way people use “hopefully” in a sentence without an explicit referent. “Hopefully” appears to be an adverb, but in fact is intended to describe the attitude of the speaker. This is usually branded as ungrammatical rather than impressionistic.

  16. Grammar should be seen as a guide, not a tyrant. Give unto grammar what is grammatical, and unto meaning what is meaningful.

  17. I should add that I myself spurn such adverb antics when speaking or writing high-flown German. I go for no-nonsense, explicit hypotaxis. Probably I have never recovered from infection by the anti-hopefully bug.

  18. People who scream about “This matter will hopefully be resolved …” don’t get upset about “It is evidently the case that …”, do they ? But the latter means “it is the evident case that …”. “Evidently” is not a way of being, but an attribute of “case”.

  19. John Cowan says:

    Indeed, there’s a theory (unfortunately without clear evidence) that sentence hopefully is a calque of hoffentlich. English has plenty of other sentence adverbs, though.

  20. That’s interesting, John. Apart from theories, though, I myself find the similarity entertaining and instructive. In German you get elegance points for sneaky “adverbs”, in English you get stomped on.

  21. Wait, I just noticed: is “sentence adverb” the technical term for “adverb that ads something other than the verb” ?

  22. Instead of “adverb” it could be called an “advert”, since it adverts to something other than the verb.

  23. John Cowan says:

    Stu: Yes, it is. Here are a number of examples that are considered unexceptionable:

    “Ideally a book would have no order to it, and the reader would have to discover his own.” (Mark Twain)

    “Ironically, women who acquire power are more likely to be criticized for it than are the men who have always had it.” (Carolyn Heilbrun)

    “Apparently, a democracy is a place where numerous elections are held at great cost without issues and with interchangeable candidates.” (Gore Vidal)

    “Certainly, travel is more than the seeing of sights; it is a change that goes on, deep and permanent, in the ideas of living.” (Miriam Beard Vagts)

  24. If those examples are considered unexceptionable by the same people who fuss about “hopefully, …”, why do they make an exception for “hopefully” ? Hopelessly this will all get straightened out one day.

  25. On the classification of adverbs, James McCawley’s system is straightforward and easy to explain (though I have reservations). First, a modifier is something which when added to a phrase gives you a a phrase of the same category. Adjectives are modifiers of nouns (actually, McCawley says N-bars). Other modifiers are called adverbs (or perhaps adverbials). You can classify various adverbs according to the category of what they modify. Among other types, there are adverbs that modify verbs (e.g. “completely”), those that modify verb phrases (e.g. the manner adverbs), those that modify sentences (e.g. “possibly”), and those that modify performatives (e.g. the “frankly” in “Frankly, my dear, I don’r give a damn.”).
    This is summed up in McCawley’s “The Syntactic Phenomena of English”.

  26. If those examples are considered unexceptionable by the same people who fuss about “hopefully, …”, why do they make an exception for “hopefully” ?

    That’s just one subquestion under the general category “Why do the peevers peeve?” If examined closely, almost all peeves are just as inconsistent and nonsensical as that one, but we humans are extraordinarily good at rationalization — it may well be what we’re best at. We can always find a way to finish sentences like “I did that because…” or “I think that because…”

  27. Well, that’s fair. I’m just not versed in the peevish details of “hopefully”. In particular I was not aware of the “sentence adverb” category, exemplified by “ideally”, “apparently” u.t.d. Before I read John’s list, I suppose I didn’t care one way or another about “hopefully” (apart from the fact that hope is inimical to my nature). Now I must take care not to fall into anti-peevery.

  28. A very minor point, but
    “I found the odd-looking quinzhee in the OED and discovered it’s pronounced /ˈkwɪnzi/ (QUIN-zee, just like the traditional pronunciation of Quincy)”

    … is there, then, a different modern way to pronounce Quincy, and what is it?

  29. John Cowan says:

    /kwɪnsi/. I bet most of the people who bear it as a first name use this spelling pronunciation, which agrees with most, if not all, other words in -cy.

  30. marie-lucie says:

    What about the English opium-eater and writer De Quincey? I read his book but never heard his name.

  31. David Marjanović says:

    What, nobody has brought up Georgian dede “mother” and mama “father” (even “church father”)?

    Also, zh in an Athabaskan language is of course [ʒ].

    I have often wondered about the Chinese words 爹/diē and 娘/niáng vs. 爸/bà and 妈/mā.

    …The first pair is wholly news to me. But the formal term for “parents”, fùmǔ (I can’t find the characters right now), was explained to me as being composed of “father” and “mother”, presumably *pa- and *ma- after a few sound changes. – The informal version is simply bàba, māma.

    I wonder if diē could be *da- and niáng could be *nganga or rather *nyanya?

    And maybe the accusation of Westernizing was just about being too informal?

    “Why do the peevers peeve?”

    For much the same reason that the heathen rage: it’s newfangled, it’s not English as she was spoke (or at least written…!) when and where they grew up. (Naturally, this includes cases of the recency illusion – of people imagining a vain thing, so to say.)

  32. John Cowan says:

    the heathen rage

    That “rage”, by the way, is a Septuagint reading; in the MT the “nations conspire”, which is a better semantic rhyme with the “peoples plot” of the following line. The KJV’s imagine meant ‘plot’, so it is misleading nowadays, remembered only in the words of the Treason Act 1351, “when a man doth compass or imagine the death of our lord the King”; both verbs mean ‘plan, plot’.

    But it’s a good wisecrack anyway.

  33. The heathen raged (in Church Slavic) last year in this post.

  34. both verbs mean ‘plan, plot’

    But still as a mental, not a speech, act. Some physical act demonstrating the compassing / imagining is needed to prove treason.

  35. John Cowan says:

    Sure. The elements of a crime are one thing, the evidence needed to prove those elements is another. In high school, one of my social-studies teachers was a lawyer, and she gave us little exercises of the form “The definition of murder is X. The following facts about Y’s actions are undisputed: [...]. Did Y commit murder?”, where the answers were “Yes”, “No”, or “Can’t tell”. It was all very good training in making (or not being able to make) inferences from given data. In the U.S., the elements are prescribed in the Constitution, inevitably so given our lack of a king after 1776.

    In the particular case of the Treason Act, the special requirements for proof were removed in 1945. The U.S. Constitution perpetuates a strengthened version of the 1695 Act’s requirement for two witnesses to the same offense, requiring two witnesses to the same overt act. Consequently, trials for treason in the U.S. are very rare, with only about 40 prosecutions in our whole history and few convictions. John Brown was hanged for treason against Virginia and Thomas Dorr got life imprisonment for treason against Rhode Island (but was released as part of a general amnesty). There have been a few other state prosecutions, including Joseph Smith’s: he was acquitted once, retried in a different state, and lynched while awaiting trial.

  36. Missed this entire thread.

    In addition to 爹/diē and 娘/niáng, and 爸/ and 妈/, there is also the formal 父亲/fùqīn and 母亲/mǔqin, straight from Japanese 父親/chichi-oya and 母親/haha-oya.

  37. David Marjanović says:

    That “rage”, by the way, is a Septuagint reading; in the MT the “nations conspire”

    Ooh! Like in that Egyptian source about the Sea Peoples, where “the foreign peoples conspired on their islands”!

    straight from Japanese

    Wow, I really didn’t expect that kind of word among Japanese loans.

Speak Your Mind