Transcript Lysbilde 1

Malé poměry, velké změny:
Language variation and change in North Saami
Laura A. Janda
Lene Antonsen Jussi Ylikoski
North Sámi…
• A Uralic (Finno-Ugric) language
• About 20,000 speakers
• In contiguous regions of northern Norway, Sweden, and
Finland
• Unique in Europe as a minority language in contact with
majority languages from two different language families:
Indo-European and Finno-Ugric (Ylikoski 2009:201-202)
• For several decades in 20th century, children were forcibly
removed from their families and forced to live at residential
schools
5 = North Sámi
3
An ongoing language change:
NPx is being replaced by ReflN
Two examples from Elle Márjá Vars’ novel Kátjá
NPx (possessive suffix, HIGH morphological complexity):
(1a) Kátjá...
ollii
latnjasis
Kátjá.NOM
reach.IND.PRET.3S room.ILL.SG.PX.3S
‘Kátjá... got to her room’
ReflN (analytic construction with reflexive genitive pronoun):
(1b) Kátjá... ollii
iežas
latnjii
Kátjá.NOM reach.IND.PRET.3S REFL.GEN.3S
‘Kátjá... got to her room’
room.ILL.SG
WALS Feature 57a: Possessive affixes
North Saami
ReflN is not itself morphologically complex
Gen/Acc reflexive pronoun is something one
needs for other purposes...
This pronoun does its expected jobs
aside from expressing possession
Pronominal use of reflexive pronoun:
Son
ii
orron
3S.NOM
NEG.IND.PRS
seem.IND.PRT.CONNEG
dovdame
iežas
speadjalis...
recognize.AKTIO.ESSIVE REFL.ACC.3S mirror.LOC.SG
“She didn’t seem to recognize herself in the mirror...”
NPx,
by contrast, is
morphologically
complex
NPx requires a large quantity of unique
morphology
•
•
•
•
•
•
•
[N. Saami has 3 types of noun stems: vowel stems, consonant stems, and
contracted stems]
NPx expands the paradigm of a noun from 10 forms to 91
– see 81 additional forms for guoibmi “partner” on handout with forms
involving NPx-unique morphology boldfaced
There are two full sets of 9 possessive suffix forms, one set attaches after
vowels and one set attaches after consonants (compare Gen/AccSg with
IllSg in handout)
NPx conditions unique changes in case endings: Illsg -i/-ii > -s-/-asa-; LocSg
-s/-is > -st-/-isttá-/-istti-; IllPl -ide/-iidda > -idas-/-iiddásWith NPx certain case endings have additional morphophonemic variants
depending on the type of stem: LocSg, Com Sg=LocPl, AccPl=GenPl, IllPl
NPx conditions additional morphophonemic alternation in the stem, e.g. i ∼
á
NPx requires insertion of the possessive suffix inside the Comitative Plural
ending
Morphological complexity in the face of
intense contact
• NPx is much more morphologically complex than ReflN
• North Saami is under intense pressure from both
Germanic (Norwegian and Swedish) and Finnish
• Many North Saami speakers have reclaimed the language
as adults
• Language contact and 2nd language learners can lead to
morphological simplification (Trudgill 2002, McWhorter
2007, Bentz & Winter 2013)
• In this situation, there may be an advantage for the
morphologically simpler form: ReflN
North Saami on Trudgill’s (2011) scale
Trudgill (2011) has a 6-point scale
1 = strongest trend toward complexification
6 = strongest trend toward simplification
North Saami receives a “4” due to:
– small size
– loose network
– high level of contact
Expectation: Reduction in paradigmatic redundancy, loss of
morphological categories compensated for by an increase in
transparent analytical structures (like ReflN)
An example of an S-curve
(cited by Blythe & Croft 2012)
S-curve
The S-curve in our Data
Our S-curve is
from this data
• Literary texts: 530,000 words, three age groups, two
geographic regions
– 2,272 examples, full analysis by hand
• The New Testament (1998): 136,522 words
– 1,530 examples, full analysis by hand
• Newspapers: 10M words from three newspapers (19972011)
– 29,964 examples of words with frequency ≥5, partial
automatic analysis, a lot of cleaning by hand
• Total: 33,633 examples
Young
Mid
The S-curve:
longitudinal
data from
literary texts,
showing only
anaphoric and
endophoric
use
Old
Frequency
Does frequency explain the distribution?
• Expectation: High frequency words are shielded from the
change, while low frequency words are vulnerable
☛ Is NPx more frequent with high frequency words?
0.6
0.4
0.2
0.0
Proportion NPx
0.8
1.0
No evidence that high frequency helps to retain NPx
News data: Pearson's correlation = -0.14, p = 0.0001, 95%
confidence interval: -0.2 -0.07
2
4
6
8
Log Transformed Word Frequency
10
CART-analysis
“Classification and regression trees and Random
forests”:
– Optimal sorting of data
– Results similar to regression, but appropriate
for non-parametric data
– Bootstrapping and measurement of variable
importance
North Saami CART analysis: variables and levels
Dependent variable: PossCon (possessive construction)
– NPx (noun with possessive suffix), ReflN (reflexive genitive pronoun)
Independent variables:
• Generation (of author)
– Old, Mid, Young
• PMClass (semantic class of possessum)
– Abstraction, Body, Event, Human, Kin, Place, Property, Other
• PMCase (case of possessum)
– Acc, Com, Ess, Gen, Ill, Loc, Nom
• PRCase (case of possessor)
– Acc, Gen, Ill, Loc, Nom, Verb
• Geography
– East, West
18
Is nominative + possessive suffix
evolving into a new Vocative case?
In preceeding statistics we excluded Exophoric Reference
Nearly all examples of Exophoric use are NPx, and most of
those are what we call “Exophoric Vocative”:
Gula, mánážan. (KP2: 6)
Note diminutive suffix
mánážan
-š > -ž intervocalically
child.DIM.NOM.SG.PX.1S
“Listen, oh my (little) child.”
This use of NPx is robust even in the Young generation
Can a possessive form become a vocative?
• The claim that a possessive construction could develop
into a vocative is not unprecedented
• Michael (2013: 157) documents the use of the First
Person Singular possessive construction with close
kinship terms, as in ina “my mother” in Nanti (spoken in
Peruvian Amazonia) as vocatives
Features of 1.sg NPx Vocative
•
•
•
•
•
•
•
Possessums are restricted to kinship terms, names, metaphorical names for
people and names or words for animals that are addressed as if they were people
Possessums are nearly always Singular (plural is not attested in our data but
known to occur rarely & in connection with religious rituals)
Possessums often have diminutive suffix and are often found in combination
with an imperative verb form
All vocatives in our literary data have 1.sg reference; in the Bible we find 1.pl in
Áhččá-met [father.NOM.SG-1SG.POSS] ‘Our Father’
Vocative meaning in N. Saami has been described since 1920s and is part of a
larger trend toward reinterpretation of NPx in Uralic languages
The development of a Vocative could be further undermining the integrity of the
NPx paradigm and thus further disadvantaging NPx vis a vis the ReflN
construction
The tendency for inflectional forms to get “recycled” into new roles when
paradigms are under pressure due to historical erosion is well documented (Lass
1990, Janda 1996)
4 Ambipositions in North Sámi
1. a. miehtá dálvvi
b. dálvvi miehtá
[over winter-G]
[winter-G over]
‘during the winter’
2. a. čađa áiggi
b. áiggi čađa
[through time-G]
[time-G through]
‘through time’
3. a. rastá joga
b. joga rastá
[across river-G]
[river-G across]
‘across the river’
4. a. maŋŋel soađi
b. soađi maŋŋel
[after war-G]
[war-G after]
‘after the war’
Ambipositions:
Adpositions that function as both prepositions and
postpositions
• Typologically rare phenomenon (Hagège 2010 )
• Such adpositions are found in Finnish, Estonian,
and the Sámi languages
• There are more of them in Sámi languages
• In Finnish and Estonian there are strong
tendencies (e.g., time with preposition, space
with postposition)
• Distribution in Sámi languages has not previously
been studied in detail
Three hypotheses
1. We expect to find regional variation in use of
ambipositions since N. Sámi is in contact with
Finnish (predominantly postpositional) in East,
and in contact with Norwegian/Swedish
(prepositional) in Central and West regions
2. We expect position to be associated with
differences in expression of meaning
3. We expect that a language with more
ambipositions will use position in a more
complex way
Ambipositions in North Sámi:
Material
• 4 relatively frequent ambipositions:
•
•
•
•
miehtá ‘over’ (2 dimensions, time and space)
čađa ‘through’ (3 dimensions, time and space)
rastá ‘across’ (1 dimension, space)
maŋŋel/maŋŋil/maŋŋá ‘after’ (1 dimension, time)
• 2 databases:
• Newspapers
• Literature
Examples from newspapers
• 1997-2011: 10M words
• Automatic extraction found 7496
sentences with the 4 ambipositions
• Goal: Tag by hand ≥100 examples for
each position for each adposition
• In all, 901 examples were tagged by
hand for types of meaning expressed
3000
Distribution of adpositions in newspaper corpus
(10M words)
2500
2000
PR
1500
PO
1000
500
0
miehtá
čađa
rastá
maŋŋel+
Examples from literature
• Examples extracted partly automatically, partly by
hand
• 652 examples of the 4 ambipositions found, all
tagged by hand
• 20 texts, three geographical regions:
• West = S. Troms (contact with Norwegian and
Swedish)
• Central = Kautokeino (contact with Norwegian)
• East = Tana (contact with Finnish)
• New translation of the New Testament
Tana
S. Troms
Kautokeino
Geographical distribution of literary texts
Distribution of adpositions in literature
300
250
200
PR
150
PO
100
50
0
miehtá
čađa
rastá
maŋŋel
Distribution of adpositions
100.00%
90.00%
80.00%
70.00%
60.00%
% PR
50.00%
% PO
40.00%
30.00%
20.00%
10.00%
0.00%
S. Troms
Kautokeino
Tana
NT
Newspapers
Distribution of adpositions
100.00%
Х2=118, df=2, p<2.2e16
Cramer’s V=0.49
90.00%
80.00%
70.00%
60.00%
% PR
50.00%
% PO
40.00%
30.00%
20.00%
10.00%
0.00%
S. Troms
Kautokeino
Tana
NT
Newspapers
A -ráigge ‘hole’ in North Saami grammar?
The advent of a prolative case marker
by Jussi Ylikoski (2014)
Words ending in -ráigge ‘hole’ have been attested in North Saami for
about 200 years (also found in S., Aanaar, and Lule Saami):
Biret livkkihii olggos duon din ođđa uksaráigge mii duoppe oidno.
‘Biret went outside through that new door that is visible over there.’
cf. Czech 7. pád: vyšla dveřmi
This would make sense since a door is a type of hole, and there
are many attestations of -ráigge with nouns that denote holes:
Máhtte oinnii Máreha láseráigge.
‘Máhtte saw Máret through the window.’
cf. Czech dívat se oknem
Cokka báddegeaži uhca ráigeráigge.
‘Stick the end of the rope through the hole.’
But it’s not limited to holes…
In a corpus of over 20M words, we find over 500 examples of -ráigge
words, and some of the most common are:
•
luoddaráigge ‘along the trail’
•
geaidnoráigge ‘along the road’
•
johkaráigge ‘along the river’
•
Deatnoráigge ‘along the Tana river’
•
eatnoráigge ‘along the main river’
•
bálggesráigge ‘along the path’
Other examples include:
•
jiekŋaráigge ‘along the ice’, luhpporáigge ‘along the steep trail’,
meahcceráigge ‘through the forest/wilderness’, šalderáigge ‘along
the bridge’, šalkaráigge ‘along the trail made of packed snow’,
guorbmevávdnageaidnoráigge ‘along the road for trucks’
So what are these -ráigge words?
✘ Are they Noun-Noun compounds?
✘ Are they adverbs?
✘ Are they nouns with a postposition?
✔ Are they nouns with a prolative case ending?
Ylikoski uses morphological, syntactic, and
semantic arguments to support the hypothesis that
-ráigge is a prolative case ending
Saami has lost a prolative case that can be
reconstructed as *-ko and we see traces in adverbials
and adpositions (cf. Czech kudy)
case ending
where
there
beside
between
close by
behind
on the side
under
before
behind
above
in
So it is natural for N. Saami to fill in the missing spot with a new prolative
Are -ráigge words Noun-Noun compounds?
Ylikoski has several arguments, but here is one:
«Eiseválddit leat láhppon, ja go láhppo, de ferte jorggihit
iežas luoddaráigge gitta dovdá fas» (Ávvir 21.2.2013: 7)
‘The authorities are lost, and when one gets lost, one has
to go back along one’s own path until one comes to a
place one recognizes.’
ReflN iežas can only refer to the path (luodda) which can
be one’s own; it can’t refer to luoddaráigge as if it were a
noun since it can’t be *one’s own along the path
Are -ráigge words adverbs?
NB: This is the preferred solution in current dictionaries and grammars
•
•
One argument is that one cannot add other adverbs to these words:
*menddo bálggesráigge ‘too along the path’ *hui láseráigge ‘very through
the window’
Another argument is that you can supply attributes, demonstratives, and
even relative clauses to these words, which one can’t do with adverbs:
‘that’
‘our’
‘good’
‘path’
‘that goes to the beach’
Are -ráigge words nouns with a
postposition?
One argument against this hypothesis is that –ráigge does not allow
metaphorical extensions, which are typical for postpositions
• Here we can compare –ráigge with postposition bokte ‘through, via’
• None of these formations are possible: *láhkaráigge ‘through the
law’, *bušeahttaráigge ‘through the budget’, *ovttasbargoráigge
‘through collaboration’, *interneahttaráigge ‘through the internet’,
*Kristusráigge ‘through Christ’, *sátnejođiheaddjeráigge ‘through the
mayor’
• But all of the above are possible combinations with postposition
bokte ‘through, via’:
lága/bušeahta/ovttasbarggu/interneahta/Kristusa/sátnejođiheaddji
bokte
Are -ráigge words nouns with a prolative
case ending?
The paradigm of bálggis ‘path’ with prolative forms
However, the prolative is certainly not a
prototypical case
• It is limited to nouns that designate
concrete objects or places that can serve
as natural paths, such as paths, rivers,
roads, fjords, doors, windows, chimneys,
and holes.
• In some examples -ráigge does behave
like a postposition
Ylikoski cites the famous words of Edward
Sapir (1921): “All grammars leak”
Filling another “hole” in North Saami grammar:
A dual for nouns: -guovttos (Ylikoski 2015)
Both pronouns and verbs have dual forms, but nouns don’t.
Here are the pronouns:
Nom.
Akk./Gen. Ill.
Lok.
Kom.
Essiiva
1sg
mun
mu
munnje
mus
muinna
munin
2sg
don
du
dutnje
dus
duinna
dunin
3sg
son
su
sutnje
sus
suinna
sunin
1du
moai
munno
munnuide
munnos
munnuin
munnon
2du
doai
dudno
dudnuide
dudnos
dudnuin
dudnon
3du
soai
sudno
sudnuide
sudnos
sudnuin
sudnon
1pl
mii
min
midjiide
mis
minguin
minin
2pl
dii
din
didjiide
dis
dinguin
dinin
3pl
sii
sin
sidjiide
sis
singuin
sinin
…and here is a verb,
the present tense of viehkat ‘run’
mun
don
son
moai
doai
soai
mii
dii
sii
viegan
viegat
viehká
vihke
viehkabeahtti
viehkaba
viehkat
viehkabehtet
vihket
Dual pronouns and verb forms
refer almost exclusively
to human beings,
usually with definite reference
But North Saami nouns,
at least as presented in
grammars, have only
singular and plural forms
Note that some Samoyedic languages have dual number for nouns
Number agreement in North Saami
SG: Bárdni lea lohkamin.
[Boy.NOM.SG is.3SG reading]
‘The boy is reading’
DU:
a. Bártnit leaba lohkamin.
[Boy.NOM.PL are.3DU reading]
PL: Bártnit leat lohkamin.
[Boy.NOM.PL are.3PL reading]
‘The boys are reading’
b. Guokte bártni leaba lohkamin.
[ Two.NOM boy.GEN.SG are.3DU
reading]
c. Bártniguovttos leaba lohkamin.
[Boy.GEN.SG-two are.3DU reading]
‘The two boys are reading’
Why -guovttos looks like a dual for nouns:
It can be used in various case forms
NOM subject: Mánáguovttos leaba skuvllas ‘The two children are at school’
GEN possessor: Mánáguoktá váhnemat leaba barggus ‘The parents of the two
children are at work’
GEN with postposition: Mánáguoktá duohken lea heasta ‘There is a horse behind the
two children’
ACC direct object: In oainne mánáguoktá ‘I don’t see the two children’
LOC possession: Mánáguoktás leat sátnegirjjit ‘The two children have dictionaries’
ILL indirect object: Oahpaheaddji attii mánáguoktái sátnegirjjiid ‘The teacher gave
the two children dictionaries’
COM: Maid mii galgat dahkat dáinna mánáguoktáin? ‘What should we do with
those two children?’
Why -guovttos looks like a dual for nouns:
It resists plural
dát amerihkálaš dánsuguovttos
‘the American dancepair’
??dát englándalaš dánsuguoktát
‘the English dancepairs’
Váhnenguovttos galgaba deaivvadit oahpaheddjiin.
‘The parent-pair should meet with the teacher’
Buot váhnemat galget deaivvadit oahpaheddjiin.
‘All parents should meet with the teacher’
??Buot váhnenguoktásat galget deaivvadit oahpaheddjiin.
‘All parent-pairs should meet with the teacher’
So what is happening in North Sámi:
• An analytic possessive construction is overtaking
synthetic possessive suffixes
• Remaining 1sg possessive suffix use is being
reinterpreted as a vocative case
• Distribution of adpositions is heavily influenced by
contact languages
• Prolative case
• Dual number for nouns