Typecraft v2.5
Jump to: navigation, search

Difference between revisions of "Annotating Runyankore-Rukiga"

 
(18 intermediate revisions by 2 users not shown)
Line 1: Line 1:
Runyankore-Rukiga refers to two of the four closely related languages of Uganda which form part of a new standardized language of Uganda together with Runyoro and Rutooro. Sometimes Runyakitara is used to refer to the standardized form of these four languages; see for example the following still incomplete [http://en.wikipedia.org/wiki/Runyakitara_language| wikipedia article ].
+
Runyankore-Rukiga refers to two of the four languages that form Runyakitara which is a new standardized language of Uganda together with Runyoro and Rutooro. Sometimes Runyakitara is used to refer to the standardized form of these four languages; see for example the following still incomplete [http://en.wikipedia.org/wiki/Runyakitara_language| Wikipedia article ].
  
Here we discuss questions related to the annotation of Runyankore-Rukiga texts and sentence collections in TypeCraft. The material is annotated by [[User:Allen| Allen Asiimwe]] who is a lecturer at Makerere University. She is a speaker of Rukiga. In addition [[User:Justus Turamyomwe|Justus Turamyomwe]] and [[User:Misah Natumanya|Misah Natumanya]] both graduate students at the Linguistic Department at the Norwegian University of Science and Technology [http://www.ntnu.no NTNU] annotate.
+
Here we discuss questions related to the annotation of Runyankore-Rukiga texts and sentence collections in TypeCraft. The material is annotated by [[User:Allen| Allen Asiimwe]] who is a lecturer at Makerere University, and a TypeCraft Advisory Board Member. She is a speaker of Rukiga. In addition [[User:Justus Turamyomwe|Justus Turamyomwe]] and [[User:Misah Natumanya|Misah Natumanya]] both former graduate students at the Linguistic Department at the Norwegian University of Science and Technology [http://www.ntnu.no NTNU] have annotate the Runyankore-Rukiga corpus on TypeCraft.  
==Annotating in TypeCraft==
+
The following example shows a standard TypeCraft annotation. TypeCraft uses three tiers for linguistic annotation: a translational tier, a gloss tier and a part-of-speech tier. When interlinear glosses (IG) are exported from TypeCraft the translational and the gloss tier are collapsed for presentational reasons. You need to have a login to TypeCraft in order to annotate. Simply write and e-mail using the mail address given on the [[http://www.typecraft.org/tc2wiki/Main_Page Main Page]]
+
  
 +
==Annotating in TypeCraft==
 +
The following example shows three tiers of a TypeCraft annotation. The TCEditor allows the annotations on several additional tiers, a free translation tier, a morph tier, a baseform tier, a meaning tier, a gloss tier and a part-of-speech tier. When interlinear glosses (IGT) are exported from TypeCraft the meaning and the gloss tier are collapsed for presentational reasons. At the time of writing You need to have a login to TypeCraft in order to annotate. Go to *login* in the upper right corner of your browser window to request a login to TypeCraft.ca
 +
this Baseform tiers cannot yet be exported.
 +
 
(1)<Phrase>9467</Phrase>
 
(1)<Phrase>9467</Phrase>
  
  
==The free morphemes ''aha'' and ''omu''==
+
====Annotating the Participle form of the verb====
 +
 
 +
The participle form of the verb is indicated by the morpheme '''-riku-'''.
 +
Although this morpheme is complex, it should not be split into -ri- and -ku- in this case, since it
 +
is the combination of these two morphemes that marks the participle as such.
 +
--[[User:Justus Turamyomwe|Justus Turamyomwe]] 11:13, 13 October 2011 (UTC)
 +
 
 +
<Phrase>18220</Phrase>
 +
 
 +
 
 +
'''Hi Justus,''' I have seen your message about  annotating and marking the participle in RR, however i have some reservations about -riku- as a marker:
 +
 
 +
1. Does it apply to all participles in RR (I.E PRESENT, PERFECT, 'ALREADY', AND 'NOT YET'? OR Is it an exclusive participle marker in RR?
 +
 
 +
2. Could it be that "-ku-" is a mere verb infinitive form used in compound structures, because "Another form of the verb often used in compound verbs is the infinitive" and -ri- is a tense marker. as suggested here; http://webspace.ship.edu/cgboer/morphology.html
 +
 
 +
3.Must we combine two morphemes each with its own independent meaning to mark participle, and in any case what kind of participle?
 +
 
 +
4. Do we have enough literature in Bantu languages to support this claim, because if it is indeed true, it should at least have some varriants or at least cognates in lanuages like Rutooro and Runyoro, LUGANDA and maybe Ruhaya
 +
--[[User:Misah Natumanya|natumanyam]] 15:27, 13 October 2011 (UTC)--[[User:Misah Natumanya|natumanyam]] 15:27, 13 October 2011 (UTC)
 +
 
 +
 
 +
'''Dear Misah''',
 +
 
 +
I have copied your interesting comment from Justus' usepage to this page, hoping that more people see it.
 +
--[[User:Dorothee Beermann|Dorothee Beermann]] 16:09, 13 October 2011 (UTC)
 +
 
 +
 
 +
'''Dear all''',
 +
 
 +
This is an interesting discussion. Yesterday, Justus
 +
sent me a mail about participles. He wanted to check whether there is a
 +
morphological boundary between "-riku-" in the word "arikuhika".
 +
 
 +
I decomposed it as a - ri - ku - hik - a (for subject pronoun, linking
 +
verb, infinitive form, root and mood respectively). The linking verb
 +
occurs in the following patterns, as examples:
 +
 
 +
(1) tarikuzayo "He/she will not go there". The morphs are ti - a -ri - ku
 +
- z- a -yo (negative, subject pronoun, linking verb, infinitive, root,
 +
mood, enclitic). There is progressive assimilation between ti- and -a-.
 +
(2) Turi omu nju "We are in the house". The morphs are Tu -ri o - mu n-ju
 +
(Subject pronoun, linking verb; preprefix, locative marker; class prefix,
 +
root respectively).
 +
 
 +
I am trying to show that the -ri- in "arikuhika" is the same in as the one
 +
in (1) and (2) above. The only exception is that when it is associated
 +
with a verb it is compounded with it. For other words (e.g. nouns and
 +
prepositions), it is written separately as in (2) above.
 +
 
 +
Therefore, there is a morphological boundary between -riku- in
 +
"arikuhika".
 +
 
 +
Thanks for letting me join the discussion.
 +
 
 +
Kind regards,
 +
 
 +
Celestino Oriikiriza
 +
 
 +
 
 +
Dear all,
 +
 
 +
I would like to draw attention to Taylor's book where he states on page 167 that '''for the participial present continuous: "the tense/aspect is marked by the infix -riku- replacing the indicative prefix ni-/n-" '''.
 +
 
 +
Justus and I have been going back from using Taylor by annotation ''riku'' as one morpheme, and decomposing the morpheme into
 +
''ri'' and ''ku'' as has been suggested above.
 +
 
 +
--[[User:Dorothee Beermann|Dorothee Beermann]] 16:01, 16 October 2011 (UTC)
 +
 
 +
 
 +
'''The Participle in Runyankore-Rukiga'''
 +
 
 +
Thank you Justus for starting this interesting debate; Misah, Celestino and Dorothee, thank you for your contributions. First I would like to understand whether participle forms mark mood or tense, or both. However, below are my submissions on the topic:
 +
1. ''-riku-'' is a present participle marker as well as the 'not-yet' only in the negative form:
 +
(a) The present participle:
 +
 
 +
<Phrase>25958</Phrase>
 +
 
 +
(b) The 'not yet' participle form
 +
 
 +
<Phrase>25961</Phrase>
 +
 
 +
 +
2. In complex sentences, it is carried by the subordinate clause.
 +
(see the illustrations above)
 +
 
 +
3. -riku- may not be decomposed further, since ''-ri-'' on its own and ''-ku-'' on alone have different meanings. I would not mark ''-ri-'' in participle verb as a linking 'verb', since the reading of 'linking' is not presented. In addition we should note that we cannot replace for instance ''-ri-'' with another morph and get a meaningful string of morphs, which shows that the two segments are morphologically inseparable(save for some sub-dialects of Runyankore where ''-ri-'' is deleted in speech). How do we analyse this? In this case, why is it that it is only -ri- that is deleted not -riku-?:
 +
 
 +
<Phrase>25969</Phrase>
 +
 
 +
4. In Rukiga, some speakers mark the present pariticiple with ''-ra-''. So, are ''-ra-'' and ''-riku-'' allormorphs or just variants?
 +
examples:
 +
 
 +
<Phrase>25964</Phrase>
 +
<Phrase>25963</Phrase>
 +
 
 +
 
 +
 
 +
'''Does -riku- always mark present participle?'''
 +
 
 +
''-riku-'' is also a relative marker in adjectival phrases:
 +
 
 +
<Phrase>25967</Phrase>
 +
 
 +
<Phrase>25966</Phrase>
 +
 
 +
Over to you!
 +
 
 +
--[[User:Allen Asiimwe|Allen Asiimwe]]--
 +
 +
 
 +
Thank you Justus
 +
====The free morphemes ''aha'' and ''omu''====
 
The free forms ''aha'' and ''omu'' are derived from the locative classes ''ha-'' and ''mu-'' via the affixation of an initial vowel. Both
 
The free forms ''aha'' and ''omu'' are derived from the locative classes ''ha-'' and ''mu-'' via the affixation of an initial vowel. Both
 
forms may occur as part of a locative expression, as determiners or as pronouns. In the example below ''aha'' is categorised as a preposition. On of the projects the Runyankore-Rukiga group is working on is to develop a better understanding of the role that both elements play in the grammar of Runyankore-Rukiga.
 
forms may occur as part of a locative expression, as determiners or as pronouns. In the example below ''aha'' is categorised as a preposition. On of the projects the Runyankore-Rukiga group is working on is to develop a better understanding of the role that both elements play in the grammar of Runyankore-Rukiga.
Line 25: Line 139:
  
  
== Is there a preposition  'of' in Runyankore-Rukiga?==
+
==== Is there a preposition  'of' in Runyankore-Rukiga?====
  
 
The following sentences taken from the TC database have in common that they contain a phrase that is translated using the English preposition ''of''. The corresponding element in RR is sometimes a free, sometimes a bound morpheme.  
 
The following sentences taken from the TC database have in common that they contain a phrase that is translated using the English preposition ''of''. The corresponding element in RR is sometimes a free, sometimes a bound morpheme.  
Line 37: Line 151:
 
<Phrase>1947</Phrase>
 
<Phrase>1947</Phrase>
  
==Diverging noun class systems for Runyankore-Rukiga==  
+
====Diverging noun class systems for Runyankore-Rukiga====
 
Morris.H.F. and B.F.R. Kirwan(1972) and Taylor, C. (1985) [[Bantu Bibliography| see Bantu Bibliography]] feature two different noun class systems.  
 
Morris.H.F. and B.F.R. Kirwan(1972) and Taylor, C. (1985) [[Bantu Bibliography| see Bantu Bibliography]] feature two different noun class systems.  
 
{| style="float:left;" border="1" cellpadding="1"  
 
{| style="float:left;" border="1" cellpadding="1"  
Line 159: Line 273:
 
|mu-
 
|mu-
 
|aka (akaato)
 
|aka (akaato)
|(o)-gu-bwa
+
|
|very big/frightening dog
+
|
 
|-
 
|-
 
|20
 
|20
 
|ku-  
 
|ku-  
 
|otu (otwato)
 
|otu (otwato)
|(a)-ga-bwa
+
|(o)-gu-bwa
|very big/frigheting dogs
+
|very big/frigheting dog
 
|-
 
|-
 
|21
 
|21
 
|gu- (ogushaija)
 
|gu- (ogushaija)
 
|otu (oturo)
 
|otu (oturo)
|wa-rugwe
+
|a-ga-bwa
|Mr. Leopard
+
|very big/frigheting dogs
|-
+
|22
+
|ga- (agashaija)
+
|obu (obuta)
+
|nya-mwojo
+
|(the woman)
+
|-
+
|23
+
|
+
|ama (amata)
+
|
+
|
+
 
|-
 
|-
 
|}
 
|}
  
 
+
[[Category:Languages|Runyankore-Rukiga]]
 
+
 
+
 
+
 
+
 
+
 
+
 
+
 
+
 
+
 
+
 
+
 
+
 
+
 
+
 
+
 
+
 
+
 
+
 
+
 
+
 
+
 
+
 
+
 
+
 
+
 
+
 
+
 
+
 
+
 
+
 
+
 
+
 
+
 
+
 
+
 
+
 
+
 
+
 
+
 
+
 
+
 
+
 
+
 
+
 
+
 
+
 
+
 
+
Are classes 22 and 23 noun class? They consist of noun prefixes that can co-occur with other noun prefixes: Wa- is prefixed on animals and birds in classes 9 and 11 e.g Warukoko Mr. Hen (class 9). Wa- was commonly used in folk tales following the African tradition of personification. Nya- on the other hand is used with class1 when the speaker does not want to reveal the name of the person being talked about but still refer back to that person as exemplified below:
+
 
+
Nya-mw-ojo ‘the boy being referred to
+
Nya-mu-kazi  ‘the woman being talked about’
+
Nya-mu-gurusi          ‘the old man being referred to’
+

Latest revision as of 21:34, 27 July 2014

Runyankore-Rukiga refers to two of the four languages that form Runyakitara which is a new standardized language of Uganda together with Runyoro and Rutooro. Sometimes Runyakitara is used to refer to the standardized form of these four languages; see for example the following still incomplete Wikipedia article .

Here we discuss questions related to the annotation of Runyankore-Rukiga texts and sentence collections in TypeCraft. The material is annotated by Allen Asiimwe who is a lecturer at Makerere University, and a TypeCraft Advisory Board Member. She is a speaker of Rukiga. In addition Justus Turamyomwe and Misah Natumanya both former graduate students at the Linguistic Department at the Norwegian University of Science and Technology NTNU have annotate the Runyankore-Rukiga corpus on TypeCraft.

Annotating in TypeCraft

The following example shows three tiers of a TypeCraft annotation. The TCEditor allows the annotations on several additional tiers, a free translation tier, a morph tier, a baseform tier, a meaning tier, a gloss tier and a part-of-speech tier. When interlinear glosses (IGT) are exported from TypeCraft the meaning and the gloss tier are collapsed for presentational reasons. At the time of writing You need to have a login to TypeCraft in order to annotate. Go to *login* in the upper right corner of your browser window to request a login to TypeCraft.ca this Baseform tiers cannot yet be exported.

(1)


Annotating the Participle form of the verb

The participle form of the verb is indicated by the morpheme -riku-. Although this morpheme is complex, it should not be split into -ri- and -ku- in this case, since it is the combination of these two morphemes that marks the participle as such. --Justus Turamyomwe 11:13, 13 October 2011 (UTC)

Mbwenu omwana ku arikuhika omu kiraasi ashanga harimu omurimo ogu omushomesa yaamutebeekanisiiza ogw’okuhandiika.
“So, when the pupil enters the class, he finds work that has been prepared for him by his teacher”
Mbwenu
mbwenu
therefore
CONJ
omwana
omwana
IV1child
N
ku
ku
when
CONJS
arikuhika
arikuhika
3SGPTCParriveFV
V
omu
omu
IVin
PREP
kiraasi
kiraasi
7class
N
ashanga
ashanga
3SGfindFV
V
harimu
harimu
16beLOC
COP
omurimo
omurimo
IV3job
N
ogu
ogu
AGRwhich
REL
omushomesa
omushomesa
IV1teacher
 
yaamutebeekanisiza
yaamutebeekanisiza
3SGPASTim3SGprepareAPPLFV
V
ogw’okuhandiika
ogw’okuhandiika
IVofIVINFwriteFV
V


Hi Justus, I have seen your message about annotating and marking the participle in RR, however i have some reservations about -riku- as a marker:

1. Does it apply to all participles in RR (I.E PRESENT, PERFECT, 'ALREADY', AND 'NOT YET'? OR Is it an exclusive participle marker in RR?

2. Could it be that "-ku-" is a mere verb infinitive form used in compound structures, because "Another form of the verb often used in compound verbs is the infinitive" and -ri- is a tense marker. as suggested here; http://webspace.ship.edu/cgboer/morphology.html

3.Must we combine two morphemes each with its own independent meaning to mark participle, and in any case what kind of participle?

4. Do we have enough literature in Bantu languages to support this claim, because if it is indeed true, it should at least have some varriants or at least cognates in lanuages like Rutooro and Runyoro, LUGANDA and maybe Ruhaya --natumanyam 15:27, 13 October 2011 (UTC)--natumanyam 15:27, 13 October 2011 (UTC)


Dear Misah,

I have copied your interesting comment from Justus' usepage to this page, hoping that more people see it. --Dorothee Beermann 16:09, 13 October 2011 (UTC)


Dear all,

This is an interesting discussion. Yesterday, Justus sent me a mail about participles. He wanted to check whether there is a morphological boundary between "-riku-" in the word "arikuhika".

I decomposed it as a - ri - ku - hik - a (for subject pronoun, linking verb, infinitive form, root and mood respectively). The linking verb occurs in the following patterns, as examples:

(1) tarikuzayo "He/she will not go there". The morphs are ti - a -ri - ku - z- a -yo (negative, subject pronoun, linking verb, infinitive, root, mood, enclitic). There is progressive assimilation between ti- and -a-. (2) Turi omu nju "We are in the house". The morphs are Tu -ri o - mu n-ju (Subject pronoun, linking verb; preprefix, locative marker; class prefix, root respectively).

I am trying to show that the -ri- in "arikuhika" is the same in as the one in (1) and (2) above. The only exception is that when it is associated with a verb it is compounded with it. For other words (e.g. nouns and prepositions), it is written separately as in (2) above.

Therefore, there is a morphological boundary between -riku- in "arikuhika".

Thanks for letting me join the discussion.

Kind regards,

Celestino Oriikiriza


Dear all,

I would like to draw attention to Taylor's book where he states on page 167 that for the participial present continuous: "the tense/aspect is marked by the infix -riku- replacing the indicative prefix ni-/n-" .

Justus and I have been going back from using Taylor by annotation riku as one morpheme, and decomposing the morpheme into ri and ku as has been suggested above.

--Dorothee Beermann 16:01, 16 October 2011 (UTC)


The Participle in Runyankore-Rukiga

Thank you Justus for starting this interesting debate; Misah, Celestino and Dorothee, thank you for your contributions. First I would like to understand whether participle forms mark mood or tense, or both. However, below are my submissions on the topic: 1. -riku- is a present participle marker as well as the 'not-yet' only in the negative form: (a) The present participle:


(b) The 'not yet' participle form



2. In complex sentences, it is carried by the subordinate clause. (see the illustrations above)

3. -riku- may not be decomposed further, since -ri- on its own and -ku- on alone have different meanings. I would not mark -ri- in participle verb as a linking 'verb', since the reading of 'linking' is not presented. In addition we should note that we cannot replace for instance -ri- with another morph and get a meaningful string of morphs, which shows that the two segments are morphologically inseparable(save for some sub-dialects of Runyankore where -ri- is deleted in speech). How do we analyse this? In this case, why is it that it is only -ri- that is deleted not -riku-?:


4. In Rukiga, some speakers mark the present pariticiple with -ra-. So, are -ra- and -riku- allormorphs or just variants? examples:




Does -riku- always mark present participle?

-riku- is also a relative marker in adjectival phrases:



Over to you!

--Allen Asiimwe--


Thank you Justus

The free morphemes aha and omu

The free forms aha and omu are derived from the locative classes ha- and mu- via the affixation of an initial vowel. Both forms may occur as part of a locative expression, as determiners or as pronouns. In the example below aha is categorised as a preposition. On of the projects the Runyankore-Rukiga group is working on is to develop a better understanding of the role that both elements play in the grammar of Runyankore-Rukiga.

The following two phrases taken from the TC-database represent one of the interesting phenomena that we came across while annotating Runyankore-Rukiga: (1)

In (1) the verb kugamba meaning speak carries the applicative suffix -ir. Interestingly it seems to be the applicative marker that forces a spatial interpretation of aha, since (2) below is ambiguous between a 'spatial interpretation' and a 'topic interpretation' of the prepositional phrase headed by aha: (3)

other examples of the use of aha and omu




Is there a preposition 'of' in Runyankore-Rukiga?

The following sentences taken from the TC database have in common that they contain a phrase that is translated using the English preposition of. The corresponding element in RR is sometimes a free, sometimes a bound morpheme.

Its forms are: we,gwe, bw, rwa, ka, za, bwa, ya. 

Following Taylor (1985) -a is the possessive marker while the variation derives from agreement. Should we call this formative a preposition?



Diverging noun class systems for Runyankore-Rukiga

Morris.H.F. and B.F.R. Kirwan(1972) and Taylor, C. (1985) see Bantu Bibliography feature two different noun class systems.

Table: Noun class systems Runyankore-Rukiga
noun class Morris & Kirwan Taylor Suggested system English translation
1 mu-(omuntu) Omu (omukazi) (o)-mu-ntu person
2 ba- (abantu) aba (abakazi) (a)-ba-ntu people
3 (e)-ki-ntu (o)-mu-ti (o)-mu-twe head
4 bi-(ebintu) emi (emiti) (e)-mi-twe heads
5 n- (ente) eri (eriino) (e)-ri-isho/(e)-i-baare eye/stone
6 n- (ente) ama (amaino) (a)-ma-isho/(a)-ma-baare eyes/stones
7 ku- (okutu) eki (ekitabo) (e)-ki-ntu thing
8 ma-(amatu) ebi (ebitabo) (e)-bi-ntu things
9 bu- (obuta) en (eka) e-n-te cow
10 ma- (amata) ama (amaka) (e)-n-te cows
11 ri- (eriino) oru (orura) (o)-ru-hu skin
12 ma- (amaino) ama (amara) (a)-ka-ti small stick
13 ru- (orushozi) oru (orushozi) orukuuto (o)-tu-ti small sticks
14 n- (enshozi) en (enshozi)/enkuuto (o)-bu-ro millet
15 ka- (akantu) oru (orushaza) (o)-ku-guru leg
16 bu- (obuntu) obu(obushaza) (a)-ha-() here (definate place)
17 tu- (oturo) aka (akagiiko) ()-ku-nu here (wider area than -ha-)
18 ha- (ahantu) obu (obugiiko) (o)-mu-() inside (a place)
19 mu- aka (akaato)
20 ku- otu (otwato) (o)-gu-bwa very big/frigheting dog
21 gu- (ogushaija) otu (oturo) a-ga-bwa very big/frigheting dogs