Articles and other missings

Discussion of natural languages, or language in general.
Post Reply
Ty185
Sanci
Sanci
Posts: 28
Joined: Sun Jun 14, 2009 4:34 pm

Articles and other missings

Post by Ty185 »

So, I was just reading up on my Lating (always a good language to know a bit about for conlangery) and I was thinking about it's daughter languages. I find it funny that it has no articles, but most, if not all, of it's daughter languages do. Why did this happen? How does a languages parent go from having missing parts to adding them along the way if they didn't exist before?

User avatar
Nesescosac
Avisaru
Avisaru
Posts: 314
Joined: Tue Jul 31, 2007 10:01 pm
Location: ʃɪkagoʊ, ɪlənoj, ju ɛs eɪ, ə˞θ
Contact:

Re: Articles and other missings

Post by Nesescosac »

Classical Latin had no articles, but Vulgar Latin, the ancestor of the Romance languages, certainly did.
I did have a bizarrely similar (to the original poster's) accident about four years ago, in which I slipped over a cookie and somehow twisted my ankle so far that it broke
What kind of cookie?
Aeetlrcreejl > Kicgan Vekei > me /ne.ses.tso.sats/

Ty185
Sanci
Sanci
Posts: 28
Joined: Sun Jun 14, 2009 4:34 pm

Re: Articles and other missings

Post by Ty185 »

Yes, but how? Vulgar Latin descended from Classical Latin, so I'm trying to figure out how it gained the articles

User avatar
Nesescosac
Avisaru
Avisaru
Posts: 314
Joined: Tue Jul 31, 2007 10:01 pm
Location: ʃɪkagoʊ, ɪlənoj, ju ɛs eɪ, ə˞θ
Contact:

Re: Articles and other missings

Post by Nesescosac »

Definite articles in Germanic and Romance languages came from forms of the word "that" that lost their potency and evolved to become markers of definiteness. Indefinite articles did the same thing with the word "one".
I did have a bizarrely similar (to the original poster's) accident about four years ago, in which I slipped over a cookie and somehow twisted my ankle so far that it broke
What kind of cookie?
Aeetlrcreejl > Kicgan Vekei > me /ne.ses.tso.sats/

User avatar
Mecislau
Avisaru
Avisaru
Posts: 491
Joined: Thu Jul 24, 2003 2:40 pm
Location: Maryland
Contact:

Re: Articles and other missings

Post by Mecislau »


cromulant
Avisaru
Avisaru
Posts: 402
Joined: Tue Jul 25, 2006 10:12 pm

Re: Articles and other missings

Post by cromulant »

As Aeetlrcreejl attests, "that" became "the" and "one" became "a." This has happened all over the world, many, many times.

Ty185
Sanci
Sanci
Posts: 28
Joined: Sun Jun 14, 2009 4:34 pm

Re: Articles and other missings

Post by Ty185 »

So when in doubt when creating languages with and without articles, use that as a go to? Also, is it possible for a language with articles to have a daughter language without them and if so, how?

User avatar
Ser
Smeric
Smeric
Posts: 1542
Joined: Sat Jul 19, 2008 1:55 am
Location: Vancouver, British Columbia / Colombie Britannique, Canada

Re: Articles and other missings

Post by Ser »

Ty185 wrote:Yes, but how? Vulgar Latin descended from Classical Latin, so I'm trying to figure out how it gained the articles
Different daughter languages took different strategies.

For count nouns, the definite article was developed in (modern) Spanish and (modern) French from the demonstrative ille 'that'. The indefinite article in Spanish and the singular forms of that of French were developed from ūnus 'one'. The plural of the indefinite in French, des, comes from dē illōs/illās 'of/from those'.*

For mass nouns, the definite in both Spanish and French comes from the same origins as above. French however developed another article for indefinite mass nouns, so-called the "partitive" article (though I must stress it's not about a "part" of the object, but rather, an unclear/indefinite amount), from dē ille 'of/from that'.

Spanish - French - (Vulgar) Latin etymology
el — le, l' — *illu
la, el — la, l' — *illa
los — les — *illos
las — les — *illas
un — un — *unu
una — une — *una
unos — xxx — *unos
unas — xxx — *unas
xxx — des — *de illos, *de illas
xxx — du, de l' — *de illu
xxx — de la, de l' — *de illa

More would have to be said about the rest of Romance languages of course... In most Sardinian dialects the definite article comes from Latin ipse '(sb) himself, (sb) herself, (sth) itself' rather than ille 'that'.
----
* Penny in his History of the Spanish Language argues that the Spanish plural indefinite articles, unos and unas, aren't articles "in any real sense, and are best considered together with other quantifiers" (p. 145), something that I agree with due to them being not compulsory as non-subjects: te traje regalos 'I brought you gifts' is as grammatical as te traje unos regalos 'I brought you some gifts'. Cf. French *je t'ai porté cadeaux and je t'ai porté des cadeaux.
Last edited by Ser on Fri Sep 02, 2011 12:32 pm, edited 4 times in total.

User avatar
Ser
Smeric
Smeric
Posts: 1542
Joined: Sat Jul 19, 2008 1:55 am
Location: Vancouver, British Columbia / Colombie Britannique, Canada

Re: Articles and other missings

Post by Ser »

Ty185 wrote:So when in doubt when creating languages with and without articles, use that as a go to?
It's just quite common at least.
Also, is it possible for a language with articles to have a daughter language without them and if so, how?
Lexification of the articles, that is, making them part of words e.g. the computer > thecomputer, sounds like a nice idea.

Ty185
Sanci
Sanci
Posts: 28
Joined: Sun Jun 14, 2009 4:34 pm

Re: Articles and other missings

Post by Ty185 »

Wow, this helped me a lot. Thank you, so much Serafin

TaylorS
Avisaru
Avisaru
Posts: 557
Joined: Sat Jul 05, 2008 1:44 pm
Location: Moorhead, MN, USA

Re: Articles and other missings

Post by TaylorS »

Ty185 wrote:So when in doubt when creating languages with and without articles, use that as a go to? Also, is it possible for a language with articles to have a daughter language without them and if so, how?
PIE's nominative masculine singular inflection -s may have originated as a definite or topicalizing article on animate nouns.

I have my English descendant conlang Mekoshan lose "a/an" all together while the "the" becomes an preposition marking direct objects. Then a reduced form of "that" becomes a new definite article. Then some prepositions (the, of, to, in, at, and for) fuse with this new definite article and create a definite article that inflects for case (think German, but more cases).

hwhatting
Smeric
Smeric
Posts: 2315
Joined: Fri Sep 13, 2002 2:49 am
Location: Bonn, Germany

Re: Articles and other missings

Post by hwhatting »

Ty185 wrote:So when in doubt when creating languages with and without articles, use that as a go to? Also, is it possible for a language with articles to have a daughter language without them and if so, how?
All cases I know where languages developed a definite article, it was derived from a demonstrative pronoun.
The only cases I know of daughter languages losing the feature of having definite articles are creoles, i.e. languages that went to a phase of losing most grammatical features of the parent language and developing new, often radically different grammatical features.

User avatar
Ser
Smeric
Smeric
Posts: 1542
Joined: Sat Jul 19, 2008 1:55 am
Location: Vancouver, British Columbia / Colombie Britannique, Canada

Re: Articles and other missings

Post by Ser »

hwhatting wrote:All cases I know where languages developed a definite article, it was derived from a demonstrative pronoun.
Then you gotta read on Sardinian (where ipse 'sb themself, sth itself' was used, as mentioned above) and Cantonese (where classifiers or measure words are often used like definite articles). :P
Serafín wrote:French however developed another article for indefinite mass nouns, so-called the "partitive" article (though I must stress it's not about a "part" of the object, but rather, an unclear/indefinite amount), from dē ille 'of/from that'.

[...]xxx — du, de l' — *de illu
xxx — de la, de l' — *de illa
Something I forgot to mention: Spanish didn't gain any particular indefinite article for mass nouns, it just uses nothing: Dame agua 'Give me water' (cf. French Donne-moi de l'eau.).

hwhatting
Smeric
Smeric
Posts: 2315
Joined: Fri Sep 13, 2002 2:49 am
Location: Bonn, Germany

Re: Articles and other missings

Post by hwhatting »

Serafín wrote:
hwhatting wrote:All cases I know where languages developed a definite article, it was derived from a demonstrative pronoun.
Then you gotta read on Sardinian (where ipse 'sb themself, sth itself' was used, as mentioned above) and Cantonese (where classifiers or measure words are often used like definite articles). :P
Latin ipse, besides meaning "himself" etc., also functioned as a demonstrative pronoun and became a pure demonstrative pronoun in later Romance languages (e.g. Spanish ese / esa / eso), so I wouldn't see it as an exception. But Cantonese looks interesting - do you have any details?

User avatar
Torco
Smeric
Smeric
Posts: 2372
Joined: Thu Aug 30, 2007 10:45 pm
Location: Santiago de Chile

Re: Articles and other missings

Post by Torco »

Neqitan gana el thread

User avatar
Ser
Smeric
Smeric
Posts: 1542
Joined: Sat Jul 19, 2008 1:55 am
Location: Vancouver, British Columbia / Colombie Britannique, Canada

Re: Articles and other missings

Post by Ser »

hwhatting wrote:But Cantonese looks interesting - do you have any details?
6.1 DEFINITENESS AND DEMONSTRATIVES

Cantonese has no articles equivalent to a or the.

[...]

To a large extent, the classifiers (6.2) perform the functions of the English articles in individuating entities. When the noun phrase is a subject or a topicalized object, the presence of a classifier denotes a definite person or object [...]:
  • Ga chē jó-jyuh go chēut-háu.
    CL car block-CONT CL exit-mouth
    'The car is blocking the exit.' (not 'A car is blocking the exit')
[...]

By contrast, a noun with a classifier following the verb may be definite or indefinite:
  • Ngóh tīngyaht wúih wán go leuhtsī.
    I tomorrow will contact CL lawyer
    'I'll contact a/the lawyer tomorrow.'
Source: MATTHEWS, Stephen & Virginia Yip. Cantonese - A Comprehensive Grammar. 1994. Routledge: London, New York. Pag. 89. (Google Books link.)
Last edited by Ser on Tue Sep 06, 2011 6:19 pm, edited 1 time in total.

hwhatting
Smeric
Smeric
Posts: 2315
Joined: Fri Sep 13, 2002 2:49 am
Location: Bonn, Germany

Re: Articles and other missings

Post by hwhatting »

@ Serafin - thanks!
The Cantonese situation seems to fall a bit between a full article and cases like in some Turkish languages, where the use of a specific case marker on a word (acc.) depends on the definiteness of the word. It's closer to an article in that it's case-inependent, OTOH, as far as I can see from the examples, the absence of the marker doesn't seem to indicate non-definiteness.

User avatar
ná'oolkiłí
Lebom
Lebom
Posts: 188
Joined: Mon Aug 24, 2009 5:23 pm

Re: Articles and other missings

Post by ná'oolkiłí »

hwhatting wrote:The only cases I know of daughter languages losing the feature of having definite articles are creoles, i.e. languages that went to a phase of losing most grammatical features of the parent language and developing new, often radically different grammatical features.
Modern Georgian no longer has articles, but Old Georgian did. In fact, in Old Georgian 3rd person pronouns, demonstratives, and articles all had the same form: as independent DPs they were pronouns; before a noun they were demonstratives; and after, articles. The loss of this feature probably had something to do with the change from the Old Georgian word order N Adj to Modern Adj N.

Old Georgian is the only attested Kartvelian language with articles, but there is evidence to suggest that case marking in the Kartvelian family may have originated from articles that fused onto the ends of nouns. Because some case endings look a lot like the extant demonstratives, it might be the case that in Proto-Kartvelian demonstratives evolved into articles, which eventually fused onto nouns and were grammaticalized as case markers. Then in Old Georgian, demonstratives became used as articles again, only for Modern Georgian to lose this feature again.



This isn't about articles, but here are two interesting ways of expressing definiteness I've come across are object marking and case. Interestingly they only apply to direct objects.

Some Bantu languages mark definite objects but not indefinite objects in the verb. If there is no overt nominal object, the marker is obligatory. For instance, from Chingoni:

Va-geni va-u-guli m-gunda.
Class.II-guests II.SUB-III.OBJ-bought Class.III-farm
"The guests bought the farm"

Vageni va-Ø-guli mgunda.
guests II.SUB-Ø-bought farm
"The guests bought a farm"

Vageni va-u-guli.
guests II.SUB-III.OBJ-bought
"The guests bought it [the Class III noun, viz the farm]"

I've extended this pattern for all 3rd person nominals marked on verbs in my conlang.

Udi (a NE Caucasian language) marks definite direct objects with the dative case (which also marks indirect objects) and indefinite DOs with the absolutive.

Iš-en tängä-Ø peškašnebe äyel-ax
man-ERG money-ABS gave child-DAT
"The man gave money to the child"

Išen tängin-ax peškašnebe äyelax
man money-DAT gave child
"The man gave the money to the child"

(The stem change in 'money' is because most nouns have different absolutive and non-absolutive stems for case formation)

merijn
Lebom
Lebom
Posts: 207
Joined: Thu Dec 21, 2006 10:36 pm
Location: Utrecht Overvecht

Re: Articles and other missings

Post by merijn »

ná'oolkiłí wrote:
Some Bantu languages mark definite objects but not indefinite objects in the verb. If there is no overt nominal object, the marker is obligatory. For instance, from Chingoni:

Va-geni va-u-guli m-gunda.
Class.II-guests II.SUB-III.OBJ-bought Class.III-farm
"The guests bought the farm"

Vageni va-Ø-guli mgunda.
guests II.SUB-Ø-bought farm
"The guests bought a farm"

Vageni va-u-guli.
guests II.SUB-III.OBJ-bought
"The guests bought it [the Class III noun, viz the farm]"
This is by and large correct, but I think I need to expand on this and qualify this somewhat. For a start, Bantu language vary quite a bit in this respect, (for instance, in Swahili inanimate objects behave the same as objects in Chingoni but animate objects trigger object agreement no matter what.) Secondly, it is often the case that while indeed indefinite objects cannot trigger object agreement, definite objects do not always trigger object agreement. An object without object agreement in such a language would still be ambiguous between an definite and indefinite interpretation. So the Zulu equivalent of the second example above izimenywa zithenge ipulazi is ambiguous between "the guests bought a farm", and "the guests bought the farm". A third point is that sometimes the translation of an object that shows object agreement is indefinite, for instance in the following sentence uwathengeleni amaqanda "Why did you bought eggs" from here (page 6). The question is if in those cases the object is also indefinite in Zulu, and the author argues that that is not the case, but I am not sure if I am convinced.

An important issue about object agreement in Bantu is that it is often more closely linked to topic/focus (topics must trigger object agreement or focused NP's cannot trigger object agreement). Topic and focus are then also related to definiteness, in fact I think the functional reason why many languages mark definiteness (that is the evolutionary force that made languages mark it) is that it makes easier for people to know what the topic and the focus is (which I think is more important that knowing if an NP is definite or not in itself).
Another thing of many Bantu languages that is linked to topic/focus and thus indirectly to definiteness is the conjoint/disjoint alternation. In the most straightforward cases the disjoint form has verb focus, but the conjoint form is used if the focus is on something else than the verb and in many cases in Bantu languages if something other than the verb has focus it must follow the verb directly, so you always know in those languages what part of the sentence is focused. I believe Kinyarwanda is such a language (I cannot seem to find the examples that I read earlier, but then again, I haven't looked really hard). In some other languages, like my beloved Zulu, the alternation is linked to focus much more indirectly. In Zulu, the disjoint form is used at the end of a VP, and the conjoint form is used when the verb is not at the end of a VP. However, if something other than the verb is focused (and sentence isn't a cleft sentence) it is always following the verb and is within the VP, thus the verb is not at the end of a VP. So you have the contrast between the conjoint ngidlala ngaphandle "we are playing outside" and the disjoint ngiyadlala ngaphandle "we are playing outside". In the first example ngaphandle remained inside the VP, where it can receive a focus interpretation, and the verb is in the conjoint form because it is not VP-final. In the second example ngaphandle has moved outside the VP, and it can no longer receive a focus interpretation and the verb is in the disjoint form, because it now is VP-final.
I know that this is not directly related to definite articles, but it is one of the coolest features of Bantu languages. Also, it is one of the few areas where the easiest way to explain the data to someone not familiar with the language is by giving the generative analysis rather than stating the (confusing) facts.

User avatar
jal
Sumerul
Sumerul
Posts: 2633
Joined: Tue Feb 06, 2007 12:03 am
Location: Netherlands
Contact:

Re: Articles and other missings

Post by jal »

hwhatting wrote:creoles, i.e. languages that went to a phase of losing most grammatical features of the parent language and developing new, often radically different grammatical features.
Err... I don't think that is completely in line with what we know about creoles, and then I'm being generous. Creoles are not languages "that went through a phase of losing grammatical features of the parent language", nor do they develop "new, often radically different" grammatical features. Creoles graft the lexicon of the "parent" (more typically called "lexifier") language onto an existing grammatical substrate, which may be radically different from that of the lexifier, but is in no way "developed" (nor is any of the grammar of the lexifier "lost", it is just not copied).


JAL

User avatar
jal
Sumerul
Sumerul
Posts: 2633
Joined: Tue Feb 06, 2007 12:03 am
Location: Netherlands
Contact:

Re: Articles and other missings

Post by jal »

merijn wrote:the easiest way to explain the data to someone not familiar with the language
"easy" is een relatief begrip hier :). But thanks for the explanation, it's quite interesting.


JAL

merijn
Lebom
Lebom
Posts: 207
Joined: Thu Dec 21, 2006 10:36 pm
Location: Utrecht Overvecht

Re: Articles and other missings

Post by merijn »

jal wrote:
hwhatting wrote:creoles, i.e. languages that went to a phase of losing most grammatical features of the parent language and developing new, often radically different grammatical features.
Err... I don't think that is completely in line with what we know about creoles, and then I'm being generous. Creoles are not languages "that went through a phase of losing grammatical features of the parent language", nor do they develop "new, often radically different" grammatical features. Creoles graft the lexicon of the "parent" (more typically called "lexifier") language onto an existing grammatical substrate, which may be radically different from that of the lexifier, but is in no way "developed" (nor is any of the grammar of the lexifier "lost", it is just not copied).


JAL
That is just one of many theories about Creolization. I always understood that creole genesis is one of the most hotly debated issues of language change, and while Wikipedia is not the be-all-and-end-all of all linguistic knowledge, it does confirm my understanding.
jal wrote:
merijn wrote:the easiest way to explain the data to someone not familiar with the language
"easy" is een relatief begrip hier :). But thanks for the explanation, it's quite interesting.


JAL
Since my post I thought of an explanation in which there are three kinds of words, words that trigger the conjoint form if they directly follow the verb, for instance question particles, let's call the conjoint-triggers (CJ's), words that don't trigger the conjoin form (let's call them nonCJ's), for instance agreeing objects, and words that sometimes trigger the conjoint form and sometimes not depending on whether it has focus or is a topic, let's call those ambiguous conjoint triggers (ACJ's). Then you could say that a verb is always in the disjoint form except if it is followed by a CJ or an ACJ that is (somewhat) focused. This however fails to explain why CJ's always precede nonCJ's, and why ACJ's always precede non-CJ's when they are focused but not when they are not focused, so i'd say that my generative explanation is still the easiest explanation of all the data. Most grammar books however start with saying that the disjoint form is used at the end of a sentence or if the verb has object agreement, and the disjoint form in all other cases and then list the myriad of exceptions to this rule, and that explanation was what I was referring to when I talked about stating the confusing facts.

hwhatting
Smeric
Smeric
Posts: 2315
Joined: Fri Sep 13, 2002 2:49 am
Location: Bonn, Germany

Re: Articles and other missings

Post by hwhatting »

jal wrote:
hwhatting wrote:creoles, i.e. languages that went to a phase of losing most grammatical features of the parent language and developing new, often radically different grammatical features.
Err... I don't think that is completely in line with what we know about creoles, and then I'm being generous. Creoles are not languages "that went through a phase of losing grammatical features of the parent language", nor do they develop "new, often radically different" grammatical features. Creoles graft the lexicon of the "parent" (more typically called "lexifier") language onto an existing grammatical substrate, which may be radically different from that of the lexifier, but is in no way "developed" (nor is any of the grammar of the lexifier "lost", it is just not copied).
As merjn said, that is just one of the theories. In any case, if you treat creoles as daughter languages of the lexifier language (which is, of course, a debatable approach), all or most of the grammatical features of the parent language are lost, whatever the exact process.
ná'oolkiłí wrote:Modern Georgian no longer has articles, but Old Georgian did. In fact, in Old Georgian 3rd person pronouns, demonstratives, and articles all had the same form: as independent DPs they were pronouns; before a noun they were demonstratives; and after, articles. The loss of this feature probably had something to do with the change from the Old Georgian word order N Adj to Modern Adj N..
What happened exactly? Did the DP used as articles disaapear, or the functional distribution (DP before noun, article after noun)?
ná'oolkiłí wrote:Old Georgian is the only attested Kartvelian language with articles, but there is evidence to suggest that case marking in the Kartvelian family may have originated from articles that fused onto the ends of nouns. Because some case endings look a lot like the extant demonstratives, it might be the case that in Proto-Kartvelian demonstratives evolved into articles, which eventually fused onto nouns and were grammaticalized as case markers. Then in Old Georgian, demonstratives became used as articles again, only for Modern Georgian to lose this feature again.
That reminds me of something I didn't remember when I posted my comment - there is another case where an article (or something like it) was lost. In parts of Balto-Slavic, adjectives are marked for definiteness by adding what was historically the relative pronoun PIE *yo- to the ending:
OCS: gora vysoka - a high mountain gora vysokaja - the high mountain
This system still works in Lithuanian. In the Slavic languages, the endings of the adjectives and the pronoun/article in many instances fused; e.g. Dat.sg.m. OCS dobru-jemu became dobru.umu, dobrumu and Modern Russian dobromu. In Slavic linguistics, the endings including the article are called "long endings", the ones without "short endings".
Now, in many Modern Slavc languages the contrast of long and short endings has either been reduced to one of inflectional classes (i.e., some word classes or even individual lexemes have either a short ending or a long ending at any specified place of the paradigm) or has changed from being a definiteness contrast to something else. In Polish, the use of historically short endings is restricted to certain adjectives and paradigm positions. E.g., the nom. sg. m of adjectives is normally the historically "long" ending -y (nowy "new"), but certain adjectives that are either used predicatively or are used pronominally have the "short" zero ending (pewien (stem pewn-) "a certain", powinien "obliged" (only used as predicative). There are similar cases in Russian, but in Russian there is also still a functional contrast between the long and short endings when adjectives are in predicative position. The unmarked form is the long form (which is also the only form that can productively be used in attributive position), while the short form implies that the state described by the predicative adjective is subjective / non-essential / temporary:
shtany korotkije - (the) trousers are short
shtany korotki - (the) trousers are too short (e.g. for the speaker or the addressee)
on schastlivyj - he's happy (in general)
on schastliv - he's happy (right now)
So, this is a case where a definitiveness contrast has been lost, although it can be argued that the long form / short form contrast was no fully fledged article, as it worked only on adjectives - the mechanism could not be used to contrast definiteness on nouns that were not accompanied by an adjective in attributive or predicative position.

Post Reply