Basque accentuation: status quaestionis and remaining issues

1
Basque accentuation status quaestionis and
remaining issues

• José Ignacio Hualde
• University of Illinois at Urbana-Champaign

2
Research on Basque accentuation Starting in
18th century Larramendi (1729) But much progress
in last 15 years or so
3

• Most of 20th century
• Ormaechea (1918) Basque has tonal accent but
  not stress accent.
• With some limitations our words of three or
  more syllables have tonal accent on their first
  and last syllables.
• Minimal pairs
• Azkue (1923) Some words are monotonic and have
  as many stress-accents as syllables others are
  ditonic and also have as many stress-accents as
  syllables, except for the last syllable which is
  atonic.
• Altube (1934) contra Azkue in Basque all
  words with the same number of syllables and
  placed in the same syntactic position are
  accented in the same manner.
• Followed by other contradictory statements of the
  facts.

4
Enormous geographic variation. Most descriptions
true (not all) , but for different dialects.
There is no Basque accent.
5
2 areas with tonal accent nowadays Northern
Bizkaia Western Navarre (Goizueta, Leitza)
6
alaba gaztea ikusi dot (Northern
Bizkaian) daughter young seen
I-have alaba gaztea ikusi dot (Southern
Bizkaian) alaba gaztea ikusi dot (Gipuzkoan and
some other Bizkaian areas) NB SB A/U
contrast (e.g. Gaminde 2000, Hualde, Elordieta,
Gaminde Smiljanic2002)
7
Northern Bizkaian Basque A/U contrast
8
Two types of words A words always bear an
accent (HL) U words bear an accent if in
immediately preverbal position and also in
isolation (which we well call sentential accent
SA). Accentless phrase-medially and in other
positions.
9
Sentential accent (SA)
10

• Understood aspects
• Nature and realization of accentual contrast
  (Azkue 1932, Basterrechea 1974-75, Jacobsen 1972)
  
• Its like Japanese ! (Hualde 1988, 1991)
• Lexical distribution of accent which words and
  classes of words and morphemes are A and which
  are U rules governing position of accent (Azkue
  1931-32 for Lekeitio)
• Position of accent in A words is contrastive
  in Gernika but not in Lekeitio/Ondarroa! (Hualde
  1988, 1991 Hualde Bilbao 1992, HEE 1994)
• Intonational patterns (HEE 1994, Elordieta
  1997, 2003, Jun Elordieta 1997, HEGS 2002, Ito
  2002, IHE 2003, Gussenhoven 2004)
• Historical aspects
• Origin of lexical A/U distinction (Jacobsen 1975,
  Hualde 1993, 1995)
• Diachronic correspondences across dialects
  (Hualde 1993, 2000, 2003)

11
Less well understood aspect Nature of SA
12
The SA rule ensures that in every sentence there
will be at least one pitch-accent.There are no
completely accentless sentences (unlike in Tokyo
Japanese) Gaur goizean gure herriko alkate
barridxá etorri da herrira today morning our
towns mayor new arrived has
town-to
SA this morning our
towns new mayor arrived to the town
13
The immediately preverbal position, where SA is
assigned, is the focus position Jonek katuá
ekarri dau Jon brought the cat/the CAT Jon-ERG
cat brought has S O V katua
Jonék ekarri dau JON brought the cat O S
V Jonék katua ekarri dau JON brought
the cat (This is unless the verb itself bears
focus, in which case it receives the sentence
accent Jonek katua ekarrí dau Jon BROUGHT the
cat )
14
Is SA in Bq the same phenomenon as nuclear accent
in other European lgs?
15

• As in Rom, Germ, etc.
• -All sentences must have at least one
  pitch-accent. Since many (most) words are
  lexically U in NBB, the existence of the SA rule
  ensures that this is always the case.
• There is a default nuclear accent position (in
  neutral sentences)
• Focalized elements attract the nuclear pitch
  accent (preverbal constituent or verb)
• Bq would appear to be similar to Germanic and
  Romance in having an accent-to-focus rule and
  also a default focus position.

16
Differently from Romance, the default focus
position is preverbal, not final, but given SOV
vs. SVO the result is default nuclear accent on
O in both lgs. Jonek katuá ekarri dau / Juan ha
traído el GAto S O V
S V O John brought the cat
17
Is Bq SA Rom/Gem NA? There are important
differences 1. NB Bq has extremely little
flexibility in assigning SA within the sentence
SA can only be assigned to the syntactic
constituent immediately preceding the verb or to
the verb itself. NB Basque Spanish Jonek katuá
ekarri dau Juan ha traido el GATO
(default) Jonék katua ekarri dau JUAN ha traido
el gato katua ekarri dau Jonék el gato lo ha
traido JUAN katua Jonék ekarri dau The only case
where SA is assigned to a different position is
when the verb itself receives SA Jonek katua
ekarrí dau Jon DID bring the cat Jonek katua
ekarrí egin dau Jon BROUGHT the cat
18
2. There are some striking cases of dissociation
between focus and SA One case that has
attracted some attention A non-phrase-final A
word within the preverbal constituent can be
given focal prominence expressed by increased
duration and a larger pitch excursion (Elordieta
Hualde 2003, Ito 2002) and by phrasing (G.
Elordieta 2003) LÉKU baltzá ikusi dot I have
seen the black PLACE place black
seen I-have
19
But a U word in the same context cannot receive
an accent KATÚ baltzá ikusi dot cat
black seen I-have Jon saw the black
CAT Even if contrastive focus, e.g. cat not
dog, is intended. Only possibility katu baltzá
ikusi dot (HEE 1993,1994, Elordieta 1997, HEGS
2002, Arregi 2004, and especially Elordieta 2003)

20
SA is assigned either a) to immediately
preverbal word (default)orb) to verbIn order
to be overtly focalizable a word must1. be
within preverbal XP or verband2. have an
accent, either lexical or SA
21
Another case Obligatory cliticization of
lexically U synthetic (finite) verbs to preceding
U word for purposes of SA assignment (HEE
1993). nire laguná etorri da my friend/MY
FRIEND has arrived my friend come
has nire laguna datór my friend/MY FRIEND
is coming, nire laguná dator Only with
synthetic verbs and when both verb and
immediately preceding word are U. Only in
declarative sentences (not in questions)
22
3. SA is assigned to the immediately preverbal
position even after a lexically accented word
with narrow focus within the focalizable
preverbal constituent. KÓLdon laguná etorri da
(Elordieta Hualde 2003)
SA KOLDOs friend has arrived SA on laguna
friend can be greatly reduced in magnitude, but
is always there. In this case focus enhancement
and SA are dissociated.
23
4. In complex sentences only one of the preverbal
positions is the locus of SA. Whether the SA is
assigned before the main or the embedded verb
depends on syntactic structure (HEE 1974, A.
Elordieta 2001, 2002, Arregi 2004). Embedded
preverbal position in complement clauses Lagunák
ekarri dabela esan dau Jonek Friend brought
has-C said has Jon Jon said that our
friend brought it Before main verb in Relative
clauses Katua ekarri daben laguná ikusi dot /
katuá cat brought has-C friend
seen I-have I saw the FRIEND who brought the
cat/ I saw the friend who brought the CAT Also
causal and time abverbial subordinate
clauses Lagunak ekarri dabelakó haserratu zan
lagunák ekarri dabelako He became angry
because our friend brought it Lagunak ekarri
dabenián haserratu zan / lagunák He became
angry when our friend brought it
24
Why?
25

• In some varieties (e.g. Ondarroa) SA is assigned
  to the penultimate of the phrase (it has been
  historically retracted). In these varieties
  lexical U/A contrast is neutralized in those
  positions where SA is assigned (words in
  isolation and in focus position), but not in,
  e.g., topic position of phrase medially
• lagunantzáko for the man (U) for the men
  (A)
• txakurre lagunantzáko ekarri dot
• dog friend-for brought have
• I brought the dog for the FRIEND/ for the
  FRIENDS
• lagunantzako txakúrre ekarri dot for the friend,
  I brought THE DOG
• lagunantzáko txakúrre ekarri dot for the
  friends, I brought THE DOG
• SA rule produces neutralization
• Contrast with synthetic verbs
• lagunantzakó da it is for the man (U SA
  assigns accent to penultimate of phrase,
  including clitic verb)
• lagunantzáko da it is for the men (A lexical
  accent on penultimate of word)

26
To conclude SA in Northern Bizkaian Basque has
some common properties with NA in other European
languages but it is a different phenomenon, only
partially related to the expression of focus. The
main function of SA is to ensure that every
utterance will surface with at least one pitch
accent. Related to the expression of focus
because it is usually assigned to an element in
focus position. But there are a number of obvious
mismatches between pragmatic focus and SA. This
is an aspect of Basque accentuation that requires
more research.
27
PART II The Western Navarrese Pitch-Accent
System Goizueta, Leitza (Western Navarre, on
Gipuzkoan border)
28
(No Transcript)
29

• Western accentual systems
• Northern Bizkaian
• 1a. Gernika-Getxo A/U contrast, initial rise
  (high plateau), U phrases 1
• 1b. Markina-Ondarroa A/U contrast, initial
  rise, U phrases 2
• 2. Southern Bizkaian
• A/U contrast, no initial rise, U phrases 2
• 3. Bergara Southern Bizkaian, but no U words.
• 4. Azkoitia 3
• 5. Central (Gipuzkoan) 2, marked 1

30
For the most part, the historical connection
among these systems is clear.
31
Lexical correspondence Northern
Bizkaian Central U 2 A 1 Explanation
progressive retraction of accentual peak and
reinterpretation of syllables bearing
non-accentual phrase-initial rises (usually the
second syllable) as accented. O O O O O
gt O O O O O
32
Northern Bizkaian
33
Mallabia (transitional dialect)
(a) lagùnana da it is the one of the friend (b)
lagúnena da it is the one of the friends
34
Mallabia In citation form, U words have a rise
followed by a slow fall. A have a rise followed
by a rapid fall
(a) frutèrue da it is the fruit seller (b)
frutérue da it is the fruit bowl
35
In non-phrase-final position, U words fail to
induce downstep of a following accent.
(a) lagùnan alàbie da it is the friends (sg)
daughter (b) lagúnen alàbie da it is the
friends (pl) daughter
36
Important confirmation that Central 2 system
derives from NB type system In NB, plurals are
A. In the Central system plurals have the marked
1 pattern, but only with short stems. This
follows directly from the distribution of
phrase-initial rises in the transitional NB area
where accents have been retracted Gernika Marki
na Central gi/xoná\ gi/zó\na gizóna the
man gi/xó\nak /gí\zonak gízonak the
men pe/lotaridxé\ pe/lotarí\xe pelótariya the
ball-player pe/lotarí\dxek pe/lotá\rixek pelótari
yak the ball-players We find marked 1
accentuation in the Central system where in the
transitional NB systems there is a rise on the
initial syllable and 2 accent everywhere else.
37
Another fact In all western systems in Bizkaia
and Gipuzkoa with sg/pl accentual distinction the
accent occurs earlier in the plural than in the
sg. (Jacobsen 1972) Lekeitio sg/pl Antzuola sg/
pl gixonák/ gixónak gizónak/gízonak
man/men ERG gixonentzakó/gixonentzáko
gizonandáko/gízonendako for In Western
Navarrese the facts have been reported to be just
the opposite the plural has final accent.
(Ormaechea 1918, 1958, Ibarra 1994, Olano
2000) Eg. Gaminde (1998234) Goizueta álabà
(sg) vs. alabák (pl) What is the correct
description/analysis of the facts? How did it
evolve? How is it related to other Western Basque
accentual systems?
38
Goizueta in collaboration with J.J. Zubiri At
first glance, system of Central type Stress
patterns in uninflected words (Zubiri 2000) a.
Regular pattern 2 até door, itxé house, lekú
place, gezúr lie, gizón man alába
daughter, txokólte chocolate, muxíka peach,
illárgi moon, atzápar paw, arráselde
evening, entsálada salad b. Marked pattern
1 áma mother, díru money, káfe coffee, éur
timber árima soul, léngusu cousin, lúmero
number, mákina machine, sóñeko dress,
kónturatu to realize
39
But Post-initial stress appears to be older in
this region than farther to the West (Larramendi
1729, Michelena 1972). The realization of
prominence (stress) is different, with phenomena
of reduction and deletion of vowels in unstressed
syllables (unattested farther to the West). Sp.
chocoláte ? txokólate gt txokólte Sp. avisádo ?
abísatu gt abístu to call
40
An important difference appears when we consider
words inflected for sg and pl. In isolation
to the mountain/ to the mountains (dat)
41
Another example
to the head/to the heads dat. /burú/
42
The final prominence in sg words in isolation is
a phrasal phenomenon. In the pl, on the other
hand, it is word-level prominence
alabari esan dio he said it to the girl
alabari eman ditet I gave it to the girls
43
alábak esan do H L the girl said it
alábák ekarri dute H H the girls brought it
Ormaechea (19186) in the part of Navarre whose
Basque I know best they delete the final accent
in singular agents in the plural, on the other
hand, they preserve it or, in other words, in
the former the final syllable has grave accent
in the latter, acute accent Gizonàk egin du.
Gizonák egin dute.
44
In other words The accent of the stem is
circumflex (HL) in the sg and acute (H) in the
pl, which has another accent on the suffix. a la
bak (sg) vs. a la bak (pl) H L
H H
45
Bisyllabic word with marked initial stress
ámak ekarri do ámák ekarri dute
the mother brought it the
mothers brought it
46

• In Goizueta, the plural accent surfaces on the
  last syllable of the word.
• alábana da
• it is the one of the daughter
• b. alábaná da
• it is the one of the daughters

47
ámana dá (sg) vs. ámaná da (pl) HL
H H it is the one of the mother
vs. it is the one of the mothers
48
Bisyllabic inflected words with unmarked 2
stress
mendîk (sg) HL HL
mendík (pl) H HL
49
Summary of observations for Goizueta - In
singular forms, stressed syllable in the stem is
associated with prominent HL (circumflex)
contour. âmak alâbak alâbantzat HL
HL HL If the
word is phase-final, there will be a second, less
prominent, peak on the final syllable of the
word mendîri HL !HL -In plural forms,
the most prominent accent is on the last syllable
of the word. The lexically stressed syllable in
the stem is usually also associated with a less
prominent peak (which does not trigger downstep
of the suffixal peak). ámâk alábâk
alábantzât mendírî H H H
H H H
HH
50
If the same syllable is the locus of lexical and
inflectional accent (in words with unmarked 2
accent with historical contraction of
heteromorphemic vowel sequence mendi-ak gt
mendik), the sg has a syllable internal dip (a L
tone) which is absent in the pl. mendîk the
mountain, ErgSg ( lt mendí ak)
HL vs. mendík the mountains, Ergpl (lt mendí
ák) H All these patterns are
consistent with a lexical contrast between HL in
the sg and H H in the pl. More like
Scandinavian, Central Franconian German/Dutch (or
Ancient Greek, see Ormaechea 1918, 1958), etc.
than like Japanese!
51
NB (Gernika) Central Goizueta ERGsg. gizonák
(U) gizónak gizónak ERGpl. gizónak
(A) gízonak gizónák
52
Sg/pl in Bizkaian and Gipuzkoan is explainable
under the assumption that sg and pl forms had
different morphological structures at the
relevant historical time. Segmental evidence in
Bizkaian alaba daughter alabea the daughter
(lt alabaa) alabaak (gt alabak) the daughters
(not alabeak)
53
Under this hypothesis plurals are assigned an
accent by the same rule that assigns accents to
compounds and clitic groups Bizkaian Pre-boundar
y accent burúandi big-headed
(compound) lagunábe the friend too (clitic
group) lagúnak the friends (plural) Not in
sg. lagunak the friend ERGsg
54
Goizueta Same historical distinction in sg vs pl
morphology, but no A/U contrast lagún-a the
friend H vs. lagún ák the friends
H H mendîk the mountain, ErgSg ( lt
mendí-ak) vs. mendík the mountains, Ergpl (lt
mendíák)
55
The End (to be continued)