Lexical ultra-conservatism

[Trebor]

In this recent thread about loanwords, Linguoboy wrote:

Some languages are extremely resistant to borrowing. Siouan languages, for instance. I think there are literally three lexical items in the entire Osage language which are known to have been borrowed.

I'd like to know more about how languages the speakers of which adhere to such lexical ultra-conservatism have been able to derive so many new terms without borrowing.

Analyses and illustrative examples are welcome, along with references to articles or books on the subject. I wish that an equivalent to describe this phenomenon along the lines of "loanword" had been coined, since one would greatly facilitate research on it ("neologism" isn't specific enough).

[linguoboy]

Some languages are extremely resistant to borrowing. Siouan languages, for instance. I think there are literally three lexical items in the entire Osage language which are known to have been borrowed.

I'd like to know more about how languages the speakers of which adhere to such lexical ultra-conservatism have been able to derive so many new terms without borrowing.

I'm not sure what there is to explain. How have languages managed to derive such large lexicons in the first place? I haven't yet seen a language community which couldn't practice such ultraconservatism. Most simply choose not to.

Although this is changing due to to pressure from Global English, Standard Chinese is still rather lexically conservative. They simply calque like mad. Take an example like 反民主主義 ("anti-democratic thought"). Each element maps to a corresponding Greco-Latin morpheme:

反 anti-
民 demo
主 crat
主義 -ism (lit. "ruling thought")

Once these equivalences are well-established (which they have been in East Asian languages since at least the Meiji Era), the derivation of new vocabulary items is practically mechanical.

In Osage, the translations tend to be less literal and more descriptive. I'll add some examples when I have my dictionary to hand.

[Bristel]

náʼoolkiłí (clock in Navajo) is a nominalized form of the verb náʼoolkił "it is moved slowly in a circle". Is that an example you're looking for of neologisms? They could have borrowed it from Spanish or English but didn't. (In fact, Navajo also has been very resistant to borrowing words)

[Salmoneus]

Closer to home, a classic example is German. Sure, there are a bunch of loanwords, particularly these days, but it's managed to produce its own words in many places where other european languages tend to use loanwords (particular loans from classical languages). Fahrrad, for instance ('ride-wheel') where other languages neologise on the basis of Latin (velocipede) or Greek (bicycle), or Fernsehen ('far-see') for television.

[Pabappa]

Navajo even goes so far as to calque various placenames, and has created new words for cities that are not calques and are not pre-existing Navajo placenames either. I only know this because I visited New Mexico once and one of the people I met told me about the Navajo names for Albuquerque and other cities, and I dont remember much of any of them except that one of the translations had the word for railroad in it, proving that the name is younger than the age of rail transport.

Actually, this might extend to proper names in general. I dont speak any Navajo, but based on browsing Wikipedia their word for the Spanish language seems to be "Naakaii", which according to Wiktionary means literally "the walkers". Actually this is probably more a word for "Mexican people" than anything, and when adding the word for language on the end it means the Spanish language.

I think, to a limited extent, there is also some calquing of proper names in Japanese. They seem to like monosyllabic words based on Chinese names, such that in their word for English, the "England" part is just 英, pronounced /ei/, which means "flower petal". And for the USA, there exists the word 米国, where 米 means "rice", therefore their word for the USA means "The Rice Country". These are based on pronunciation, however, so are not actually calques. (Who among us is aware of the original meaning 2000 years ago of the word "America" anyway?)

[Trebor]

I'm not sure what there is to explain. How have languages managed to derive such large lexicons in the first place? I haven't yet seen a language community which couldn't practice such ultraconservatism. Most simply choose not to.

Sure. But I'm not fluent in any languages which are like Osage in their serious paucity of borrowings. As a native speaker of English, which has gained so many technical terms by throwing old Greek and Latin roots together, this concept is entirely outside of my experience. Plus, it's only with the industrial, scientific, and technological developments of the past few centuries that we have had to find and invent large amounts of vocabulary for "telephone", "telegram", "electrical outlet", "vacuum cleaner", "anaesthesia", "vaccine", "phoneme", "morphosyntactic alignment".

Although this is changing due to to pressure from Global English, Standard Chinese is still rather lexically conservative. They simply calque like mad. Take an example like 反民主主義 ("anti-democratic thought"). Each element maps to a corresponding Greco-Latin morpheme:

反 anti-
民 demo
主 crat
主義 -ism (lit. "ruling thought")

Once these equivalences are well-established (which they have been in East Asian languages since at least the Meiji Era), the derivation of new vocabulary items is practically mechanical.

In Osage, the translations tend to be less literal and more descriptive. I'll add some examples when I have my dictionary to hand.

It makes sense that Mandarin would be able to calque fairly easily, since the language has a long literary, political, social, and cultural history, and China's upper classes may have had access to the materials necessary to translate literally from Latin/Greek. Native American languages aren't comparable in those ways, though. I'm definitely interested in seeing those examples from your Osage dictionary.

[zompist]

I think, to a limited extent, there is also some calquing of proper names in Japanese. They seem to like monosyllabic words based on Chinese names, such that in their word for English, the "England" part is just 英, pronounced /ei/, which means "flower petal". And for the USA, there exists the word 米国, where 米 means "rice", therefore their word for the USA means "The Rice Country". These are based on pronunciation, however, so are not actually calques. (Who among us is aware of the original meaning 2000 years ago of the word "America" anyway?)

The phonetic match is closer in Mandarin: 英 is yīng (and means 'outstanding, heroic').

'America' in Chinese is 美国 měiguó 'beautiful country'. Why the Japanese use Beikoku must be an interesting story. 米 in Mandarin is mǐ.

Similarly Germany gets 德 dé 'virtuous', representing Deutsch; and France gets 法 fǎ 'law'.

But after that it's all wordy transcriptions— e.g. Xībānyá 'Spain', Mòxīgē 'Mexico'.

[----]

Actually, this might extend to proper names in general. I dont speak any Navajo, but based on browsing Wikipedia their word for the Spanish language seems to be "Naakaii", which according to Wiktionary means literally "the walkers". Actually this is probably more a word for "Mexican people" than anything, and when adding the word for language on the end it means the Spanish language.

A few proper names are borrowed. But the names for most of the 'big players' in Europe are neologisms or calques.
Naakaii does in fact translate to "the walkers" but it pretty much describes all Hispanic people, as it was applied to Spanish colonists first. I don't know why this was the name they gave them.

beyond proper names there are essentially no borrowed words. the only two I can think of are bilagáana 'white person' < americano and bilasáana < manzana

[linguoboy]

But after that it's all wordy transcriptions— e.g. Xībānyá 'Spain', Mòxīgē 'Mexico'.

The previous examples are actually abbreviations of wordier transcriptions. 美国 is a short form of 亞美利加, 法国 of 法兰西, etc. And there are other abbreviations in common use as well, such as 澳洲 for Australia (澳大利亚), 意国 for Italy (意大利), and so forth. IME, any country with a significant Chinese diaspora population tends to get shortened in this way.

[CatDoom]

(Who among us is aware of the original meaning 2000 years ago of the word "America" anyway?)

Well, "America" wasn't even a word 2,000 years ago. The name is a Latinate derivation from the first name of Amerigo Vespucci, a Florentine navigator. A couple of letters allegedly written by him and published during his lifetime greatly exaggerated his role in the early exploration of the continents, and these same letters became the medium through which most of Europe learned about the newly discovered lands. Shortly thereafter, when the German Cartographer Martin Waldseemüller produced the first map showing the Americas as separate from Asia, he named them after Amerigo.

If you want to go further, Amerigo originated as a byname for Enrico, which in turn comes from the Germanic Haimric, a compound of words meaning "home" and "ruler" or "king."

[Clearsand]

I remember reading somewhere about some inuit language that was so against borrowing that they names parts of a car after the parts of animals. For example: eyes/headlights, heart/engine, ears/mirrors, and so on.

[linguoboy]

It makes sense that Mandarin would be able to calque fairly easily, since the language has a long literary, political, social, and cultural history, and China's upper classes may have had access to the materials necessary to translate literally from Latin/Greek. Native American languages aren't comparable in those ways, though. I'm definitely interested in seeing those examples from your Osage dictionary.

By the same token, though, Standard Chinese has more of a need to coin words en masse because they need to translate technical manuals and the like. For an oral language like Osage, one only needs to coin words and expressions for everyday things.

For items like plants, animals, and food, the Osage generally did the same thing as the settlers: They extended native words and qualified them as necessary. For instance, too designated a large tuber, generally the water chinquapin (Nelumbo lutea), which was a staple crop. Potatoes came to fill a similar niche in the Osage diet, so they were called tóoska "white tubers". Sweet potatoes were called tooscée "long tubers" or tóoskue "sweet tubers". Another staple tuber, breadroot or prairie turnip (Psoralea esculenta), gave its name to beets (tóole žúe "red breadroot"), carrots (tóoleži "brown breadroot"), and turnips (tóolehtąą "big breadroot"). It's worth mentioning that the combining forms of these modifiers are often shortened; the full forms of the last three are žúuce, žíhi, and htą́ka.

One of my favourite combining elements is mą́ze which means "metal". It can be prefixed to a wide variety of terms to specify not just artefacts of metal but anything mechanical or electrical, e.g. mą́ze įįštóolą "spectacles" (įįštóolą is literally "put inside" [óolą] the eyes [įįštá]"), mą́ze oohǫ́ "oven" (lit. "metal cook"). Mą́zeie ("metal talks") is a radio and mą́zeie ówatǫe is a television set. Ówatǫe or ówatoį is a morphologically complex form consisting of the verb tǫ́pe "watch" with the valence-reducing prefix wa- and the locative prefix ó-, i.e. "one watches things in it". Nominalised, it can mean "show", "exhibition", "film", etc.

That's another very productive source of neologisms: nominalisations. Their range can be a bit bewildering. Óolą "put in" (as in mą́ze įįštóolą) is used to mean "pie" (e.g. htóolą "meat pie; sandwich", hkąącóolą "fruit pie; pie"). Oožú "pour, put in" (óožu with valence-reducer wa-) covers a slew of receptacles from bottles to pockets, e.g. níioožú "pitcher", niixócoožu "ashtray", óožuhaa "bag" ("put things into skin"). Some, as you would expect, are very idiomatic: Hcíwažu, literally "they put stuff in the house", is the word for "headright", an annual payment allotted to eligible households.

And since someone else has already mentioned ethnonyms, most of these are also derived rather than borrowed in Osage. The first White people the Osage encountred were called įįštáxį "light eyes". Coincidentally, they were mainly French-speaking. When they encountred the English, they distinguished them as mą́ąhį htą́ka "big knives", a name they eventually transferred to the Americans. Later it was restricted to army officers or government agents, eventually becoming (via metonomy) the usual word for "government". The Navajo are the haxį́ležekáaɣe (lit. "he makes spotted/speckled blankets"), the Sioux are the hpą́paxǫ ("he cuts heads"), and the Comanche are the hpá tóohka ("wet noses"). Some names are borrowed, however, such as šálaki (< Tsalagi "Cherokee") and--exceptionally--íšpaðǫ "Spanish-speaker, Mexican".

[Astraios]

Lakota has some nice neologisms too. The one everyone seems to know is šúŋkawakȟaŋ "horse" ("strange dog"). Wakȟáŋ by itself means "spiritual, supernatural, sacred, holy, tapu, etc.", but it's also used for "unexplainable, incomprehensible, mysterious", and that meaning is taken up in the oldest white-introduced neologisms like mázawakȟaŋ "gun" ("strange iron") and mníwakȟaŋ "alcohol" ("strange water").

Máza is also used to denote something mechanical or metallic as in Osage - one of my favourites is mázathuswečha "helicopter" ("metal dragonfly"), but there's also less colourful things like maswógnaka "tin can" ("he puts things in metal") and ištámaza "spectacles" ("eye metal").

Another useful derivational element is the instrumental prefixes like na- "with the feet; by supernatural power" and others, that essentially denote how the action is performed, so you can get neologisms like nat'é "he got ran over by a car" ("he died [by action of feet(/wheels)]"), tȟabnákȟapapi "football" ("they send balls flying [by kicking]"), wičhítenaškaŋškaŋ "television" ("human faces are moving back and forth [by supernatural power]"), paílepi "flashlight" ("they set it alight [by pushing (a button)]"). This is productive in normal language too, I saw recently the adverb čhunákȟaŋkȟaŋ to describe somebody who had walked in wet grass and left a trail "by knocking the dew off at every step", so it's really just begging to be exploited for neologisms.

Besides them, there's applicatives like i- "by means of it", o- "place where": wazíyatislolye "compass" ("he knows north by means of it"), owáyawa "school" ("place where he reads things"), onážaža "laundry" ("place where it gets washed by supernatural power"). And of course there's extension (and calquing) of meaning, like with wakȟáŋgli "electricity" < "lightning" ("the spirits have come back").

Some other fun neologisms: kiŋyékhiyapi "plane" ("it is made to fly"), wakȟályapi "coffee" ("things are brewed"), waípahiŋte "snow plough" ("it sweeps aside snow by pushing"), míyoglas'iŋ "mirror" ("place where he cranes his neck at himself in the water"), phutéwokič'u "elephant" ("it gives food to its own (young) with its nose").

As for borrowings from European languages, all I'm aware of are Spayóla "Mexico", pusíla "domestic kitten", and bébela "baby". Other people have derived names, except the Pawnee Sčíli and maybe a couple of others I forget. Navajo is Šináglegleǧa ("broad-striped blankets"), Inuk is Čhaȟ'óthila ("lives in a little ice house"), French is Wašíčikčeka ("ordinary whiteman"), and so on.

I remember reading somewhere about some inuit language that was so against borrowing that they names parts of a car after the parts of animals. For example: eyes/headlights, heart/engine, ears/mirrors, and so on.

Lakota does that for headlights at least: iyéčhiŋkiŋyaŋkišta ("car eyes").

[Miekko]

(Who among us is aware of the original meaning 2000 years ago of the word "America" anyway?)

Well, "America" wasn't even a word 2,000 years ago. The name is a Latinate derivation from the first name of Amerigo Vespucci, a Florentine navigator. A couple of letters allegedly written by him and published during his lifetime greatly exaggerated his role in the early exploration of the continents, and these same letters became the medium through which most of Europe learned about the newly discovered lands. Shortly thereafter, when the German Cartographer Martin Waldseemüller produced the first map showing the Americas as separate from Asia, he named them after Amerigo.

If you want to go further, Amerigo originated as a byname for Enrico, which in turn comes from the Germanic Haimric, a compound of words meaning "home" and "ruler" or "king."

I think this was the point Publipis was making.

[linguoboy]

paílepi "flashlight" ("they set it alight [by pushing (a button)]")

Instead of using the cognate prefix paa- here, Osage prefers ðuu- "do by hand", i.e. ðuuhpéece ("create fire by hand"). Similarly, ðuuhkáamą ("ring by hand") "doorbell".

wakȟályapi "coffee"

mąhkása ("black medicine"), another favourite coinage.

And I forgot to mention "clock" above. It is mį́įoðaake ("sun tells it").

[----]

Crow can be pretty dry when it comes to this stuff. Christmas is báalaammaapbaahilinneetisaa which means 'the big day in winter when nobody works'. Thanksgiving is dakáakannuusuua 'when they eat a bird'. coffee is just bilishpíta 'black water'

There are also really weird ones like shíipaachi 'thing that looks like intestines' which means...banana. It reminds me of another rather distasteful one from Cheyenne. The word for rice in that language is he'éhesono which means 'little maggots/worms'.

[Matrix]

báalaammaapbaahilinneetisaa which means 'the big day in winter when nobody works'

Hahaha that is not dry at all, that is a hilariously oversized word that is too specific in the best way.

[Astraios]

Instead of using the cognate prefix paa- here, Osage prefers ðuu- "do by hand", i.e. ðuuhpéece ("create fire by hand"). Similarly, ðuuhkáamą ("ring by hand") "doorbell".

Fun, yu- in Lakota can be used for any hand action in a general way and also for pressing buttons, but pa- means specifically by pushing or pressing something, so. And "clock" is mázaškaŋškaŋ ("metal moves back and forth" (like a pendulum clock)). Other ones you said are owášpaŋye "oven" ("place where he bakes things"), and ikȟáŋčhola "radio" ("without its rope/tie" (i.e. wireless)), mniógnake "pitcher" ("he puts water in it"), čhaȟólognake "ashtray", wóžuha "bag" (yay exact cognate). A funny one is tȟuŋkášilawičhuŋyaŋpi "the US government" ("they are our grandfathers/leaders"). Americans are Mílahaŋska ("Longknives").

Banana is zíškopela ("little yellow curve"), which is cute.

[linguoboy]

Banana is zíškopela ("little yellow curve"), which is cute.

Osage htóžąke is an interesting case. Originally, it meant "pawpaw". But according to Quintero's fieldwork, that meaning had become obsolete and it was only used for "banana". The term one of her informants gave for "pawpaw" was htóžąke eekǫ́ "like a banana"(!).

In Osage, paa- means "by pushing or pressing", not necessarily specifically with the hand. E.g. paaléke "hatch" (lit. "shatter by pressing").

[Astraios]

In Osage, paa- means "by pushing or pressing", not necessarily specifically with the hand. E.g. paaléke "hatch" (lit. "shatter by pressing").

I think pa- used to be restricted to "by pressing/squashing" generally, and pu- meant specifically "by pressing/squashing between the hands", but pu-'s become unproductive and its meaning got mostly taken up by pa-. C.f. paškíče "he extracts liquid from it [by pressure] (e.g. with a press)" vs. puškíče "he extracts liquid from it [by pressure from the hands] (e.g. by squeezing between the fingers)". At any rate, pu-, pa-, and yu- "by (any) hand action" are all a bit fuzzy.

[----]

Here's a fun one: Certain animal terminology in Crow is all messed up. The original word for dog is bishká, and it still retains this meaning. When horses were introduced to N. America, its possessive form isaashká changed its meaning from "his dog" to "his horse". Then the term for a possessed dog was derived from this: isaashkakaáshi "his proper horse > his dog". Well, there's still not a term for an unpossessed horse, so the word iichíili 'elk' was used for that. Then the same suffix from earlier was applied to make the new word for actual elks: iichíilikaashi. This is really strange: the terms for two different animals converged to refer to a single new animal!

[linguoboy]

There's a good article which I don't have the citation for which surveys North American Indian terms for "horse". In general languages either generalised an existing term for "dog" or for "deer/wapiti" depending on whether the speakers perceived domesticity or size as their most salient feature. Crow is the only language I've seen do both. Sometimes these terms were qualified (as in Lakota), but often retronyms were introduced to cover their original meanings. In Kaw, for instance, shóⁿge originally means "dog". (Cf. Osage šǫ́ke "dog; wolf", Lakota šúŋka.) Nowadays its usual meaning is "horse" and "dog" is shóⁿge óyudá[*] or shóhiⁿga.

Osage is an exception, however: hkáwa (from Spanish caballo) is one of those exceptional few loanwords I originally mentioned.

As for the Osage instrumental prefix pu-, Quintero glosses it as "by pressing down on with movement back and forth(?); by smoothing". E.g. puštáha "iron" (lit. "smooth by pressing"). It doesn't appear very productive. In this case, for example, there is a variant with ðuu- which is identical in meaning.


[*] Unfortunately, the Kaw dictionary I have doesn't gloss óyudá and I can't hazard a guess.

[R.Rusanov]

...Lakota šúŋka...

Any Lakotans coming up to Bulgaria would be quite shocked, maybe, to hear we regularly eat shunka (meaning sausage)

[Astraios]

Nope, dogs were commonly used for food after they weren't needed anymore as pack animals (and presumably before as well).

[TaylorS]

This thread makes me sad that there aren't very many polysynthetic languages that are official languages of developing countries, because these neologisms are really fun!