The Causative/Inchoative Alternation, and the Decomposition of Little v

The Causative/Inchoative Alternation, and the Decomposition of Little v
©Greg Key
Coyote Papers 19 (2012)
University of Arizona Linguistics Circle
Tucson, AZ, U.S.A.
The Causative/Inchoative Alternation, and
the Decomposition of Little v
Greg Key
University of Arizona
[email protected]
Abstract
Morphological evidence in causative/inchoative pairs in Turkish is
analyzed to determine the derivational relationship between the
transitive and intransitive members of the pairs. Three patterns are
found: 1. The transitive member is derived from the intransitive; 2.
The intransitive is derived from the transitive; 3. Both members are
independently derived from a common base. For a complete
explanation of the data, it is proposed that the verbalizing head
little v decomposes into a verbalizer (little v proper) and a discrete
‘flavor’ morpheme (CAUSE, BECOME, etc.).
1
Introduction
Transitive/inchoative alternations are pairs of verbs, one transitive and the other
intransitive, typically denoting a change of state (Haspelmath 1993). In English, this
class is mainly represented by labile alternations such as melt/melt and shrink/shrink,
where the transitive and intransitive members do not differ phonologically.
The nature of these alternations—the derivational relationship between the two
members of a pair—is a matter of some controversy. In broad terms there are three
Key, p.2
2 MORPHOLOGICAL EVIDENCE
logical possibilities: 1. The inchoative is basic, and the transitive is derived from it
(causativization); 2. The transitive verb is basic, and the inchoative is derived from it
(decausativization); 3. Neither is derived from the other, but both are derived
independently from a common base.
Hypothesis 1 (causativization) is isometric to standard semantic representations
of the transitive alternant.
(1)
CAUSE (x, [BECOME [STATE (y)]])
(2)
λx λy λs λe [∃v [CAUSE (v,e) ∧ EFFECTOR (v,y) ∧ BECOME (e,s) ∧ THEME (s,x) ∧ ϕ
(s)]]
ϕ = “random stative predicate”
(Kearns 2000)
(Koontz-Garboden 2009)
In such a structure, the causative (CAUSE) embeds an inchoative layer (BECOME). On
hypothesis 2 (decausativization), the transitive form undergoes a lexical argumentreducing operation (Chierchia (2004), Horvath & Siloni (2009)). Since the causative is
basic, it does not embed an inchoative layer. On hypothesis 3 (independent derivation),
morphemes deriving the alternants independently attach to a common base. In the
present study, morphological evidence for Turkish is analyzed to determine whether it
supports one or more of the above hypotheses.
2.
Morphological evidence
One of the key pieces of evidence in the decausativization hypothesis is the fact
that in the majority of alternations in Italian and other Romance languages, the
unaccusative alternant differs from the transitive in having reflexive morphology. This
is taken as an indication that the inchoative is derived from the transitive (Chierchia
2004). The remaining alternations in Romance are labile, as are the majority in
English.
Both morphological patterns are explained by Chierchia’s (2004) reflexive
closure operation, wherein an external event argument is identified with the internal
Key, p.3
2 MORPHOLOGICAL EVIDENCE
event argument. Exponence may be through movement (producing a labile alternation)
or reflexive morphology.
(3)
transitive
intransitive
English: break
break
Italian:
rompersi
2.1
rompere
Evidence from Middle Eastern Languages
Morphological evidence in Romance and English, both European languages,
supports or is consistent with decausativization. Middle Eastern languages provide a
different array of data. Laks (2011), who also believes decausativization to be the
universal derivational direction in the causative/inchoative alternation, finds that
while the majority of alternations in Hebrew either have a morphologically simplex
transitive and a complex inchoative (expected as a reflex of an argument-reducing
operation) or two simplex alternants, a small number of alternations display the
unexpected pattern of complex transitive/simplex inchoative, which looks on the face
of it like causativization.
(4)
a.
b.
c.
d.
simplex transitive
complex intransitive
‘wrinkle’
‘become wrinkled’
kimet
hitkamet
complex transitive complex intransitive
hirgiz
hitragez
‘make mad’
‘get mad’
simeax
samax
‘make happy’
‘become happy’
simplex transitive
simplex intransitive
zero morphology (labile alternation)
hivri
hivri
‘make healthy’
‘become healthy’
Key, p.4
e.
2 MORPHOLOGICAL EVIDENCE
complex transitive simplex intransitive
hikpi
kafa
‘freeze’
‘become frozen’
Laks (2011)
(4e) is the problematic pattern on the decausativization hypothesis, since the transitive
hikpi is morphologically complex while intransitive kafa is simplex, at least in terms of
overt morphology. Similar facts obtain in Palestenian Arabic.
(5)
a.
b.
c.
simplex transitive
complex intransitive
kasar
inkasar
‘break’
‘become broken’
sakkar
tsakkar
‘close’
‘become closed’
complex transitive complex intransitive
complex transitive simplex intransitive
qawwa
‘make strong’
qiwi
‘become strong’
(Laks 2011)
Here again, the complex transitive/simplex intransitive pattern (5c) is problematic for
decausativization. Laks, endorsing decausativization, explains the anomaly by positing
that the deviant transitives are the product of a process he calls morphological filling of
frozen lexical entries (MOFFLE). Part of the support for this is that the number of such
alternations in Hebrew is conspicuously small (44 out of 388 (11%) in his sample).
Morphological support for deriving the alternation through decausativization as a
component of Universal Grammar can be maintained as long as apparently contrary
evidence remains scant, as in the Hebrew case, and becomes correspondingly more
difficult as such evidence increases.
The Persian facts present a somewhat different picture. In the overwhelming
majority of cases, the inchoative/causative pair is realized as a light verb meaning ‘do’
and a light verb meaning ‘become’ alternating on a non-verbal base, typically a
nominal or adjective (Folli, Harley & Karimi 2004).
Key, p.5
(6)
âb
2 MORPHOLOGICAL EVIDENCE
kar-dan
âb
šo-dan
water do-ıNF
water become-ıNF
‘melt (transitve)’
‘melt (intransitive)’
Labile alternations are exceedingly rare, though the pattern does occur.
(7)
šekas-tan
šekas-tan
break-ıNF
break-ıNF
‘break (tr.)’
‘break (int.)’
In addition, there are some cases where a simplex verbal inchoative has a transitive
alternant with an affixal causative.
(8)
suz-ân-dan
sux-tan
burn-CAUSE-ıNF
burn-ıNF
‘burn (tr.)’
‘burn (int.)’
(8), with complex transitive/simplex intransitive, is the problematic pattern for the
decausativization hypothesis. Decausativization, with a simplex transitive and complex
intransitive, is the rarest pattern in Persian.
(9)
bori-dan
bori-de šod-an
‘cut (tr.)’
‘get cut (int.)’
cut-INF
cut-PRT become-INF
The presentation and discussion of evidence from Turkish constitutes the
remainder of the paper.
Key, p.6
3
3 MORPHOLOGY OF THE CAUSATIVE/INCHOATIVE ALTERNATION
The morphology of the causative/inchoative alternation
in Turkish
3.1
Data
This verbs in this study were collected from A Dictionary of Turkish Verbs in
Context and by Theme, Ralph Jaeckel and Gülnur Doğanata Erciyeş (2006), a dictionary
of over 1,000 verbs. For inclusion in the data set, the following criteria must be met: A
verb must participate in a transitive/intransitive alternation; the intransitive member of
the alternation must have a telic reading; the intransitive member must have a nonpassive reading. In cases where only one member occurs in the dictionary, the
alternation was considered to be represented, and the missing member was
independently verified and supplied. (On the data sheet, members not occurring in
Jaeckel and Erciyeş are italicized.) A total of 146 pairs comprising 292 verbs were
collected (Appendix).
3.2
Analysis
Groups A and B (16.4%) support Decausativization: simplex transitive, complex
intrasitive. Groups C through G (65%) support Causativization: complex transitive,
simplex intransitive. (This is the problem pattern for decausativization). Groups H
through J (18.5%) support Independent Alternation: complex/complex, (possibly one
case of simplex/simplex)
However, if there are null morphemes realizing little v, the actual patterns may
be different. Perhaps independent alternation can account for all the data. If either the
causative or the inchoative morpheme can be null, then all alternants are complex.
Verbalizing heads, ‘little v’s of different ‘flavors’ (Folli & Harley 2005) alternate on a
root (Harley 2006).
Key, p.7
3 MORPHOLOGY OF THE CAUSATIVE/INCHOATIVE ALTERNATION
(10)
vP
√P
DP
vP
vCAUSE
√
√P
DP
vBECOME
√
One of the more salient effects of root adjacency is allomorphy. Turkish causative
morphemes have the following forms (Göksel & Kerslake 2005):1
(11)
a. Productive (‘syntactic’) causative
-DIr following a monosyllabic stem, or a polysyllabic stem ending in a non-liquid
consonant
-t following a polysyllabic stem ending in a vowel or a liquid
b. Unproductive (‘lexical’) causative: unpredictable
-Ir, -Ar, -Art, -It
The regular or Elsewhere form occurs on stems not specified for a ‘lexical’ causative—
in other words, where it is not blocked by a lexical causative. The productive causative
(-DIr/-t) is the Elsewhere form. It is used to causativize virtually all unergative and
transitive verbs.
(12)
a.
b.
çalış-
‘work’
çalış-tır-
‘cause to work’
taşı-
‘carry’
taşı-t-
‘cause to carry’
The Elsewhere form also occurs in the causative/inchoative alternation with a root that
has no lexically specified irregular causative.
1
Capital letters indicate segments whose realization varies according to voicing assimilation for consonants,
and vowel harmony for vowels.
Key, p.8
(13)
a.
b.
3 MORPHOLOGY OF THE CAUSATIVE/INCHOATIVE ALTERNATION
don-
‘freeze (int.)’
don-dur-
‘freeze (tr.)’
büyü-
‘grow (int.)’
büyü-t-
‘grow (tr.)’
group E1
group F5
If a root has a specified ‘lexical’ causative, the Elsewhere form is blocked.
(14)
a.
düş-
c.
*düş-tür-
b.
düş-ür-
‘fall’
‘make fall’
group C
This brings up a puzzle: If causatives (and inchoative morphemes) are root-adjacent,
why isn’t there more allomorphy? Only fifteen (10.3%) of the transitive members have
‘lexical’ causatives. The total number of such verbs in all of Turkish does not exceed 30.
If all causatives are root adjacent, a much higher proportion would be expected to
show allomorphy. By way of comparison, in Jacobsen’s (1992) study of Japanese (cited
in Harley 2006), well over 300 such verbs are catalogued.
A more detailed analysis of the verbal forms may shed some light on this issue.
There is an asymmetry in the distribution of stem-final consonants. Monosyllabic stems
end in any of the following consonants.
(15)
consonant
ç [ʧ]
example
aç-
ğ (= vowel length) eğk
burk-
l
böl-
n
din-
p
sap-
r
kır-
s
sus-
t
bit-
m
ş [ʃ]
göm-
şiş-
Key, p.9
3 MORPHOLOGY OF THE CAUSATIVE/INCHOATIVE ALTERNATION
(Monosyllabic Turkish verb stems may also end in D, v, and y, though
these do not occur in the data set, for a total of 14)
In contrast, polysyllabic stems end in only a subset of these consonants: k, l, n, r,
ş, and t.
(16)
consonant
example
k
acık-
n
ısın-
r
delir-
ş
değiş-
l
yanıl-
t
dağıt-
(Even outside the data set, polysyllabic verb stems never end in any other
consonant, so the total is 6.)
Hypothesis: Monosyllables are acategorial roots. Polysyllables are roots plus one
or more functional morphemes. These morphemes compose a small, closed
morphological class, so it is expected that they would not display the full consonant
inventory of the language. In other words, monosyllabic stems are acategorial roots
plus a null functional morpheme. Polysyllabic stems are roots plus overt functional
morphemes. This would mean that none of the stems taking the –t allomorph of the
Elsewhere form are roots (groups F-G). Recall that the Elsewhere form is –t after
polysyllabic stems ending in a vowel or a liquid (and –DIr everywhere else). In that
case, what is the morpheme between the root and –t?
Note that the sequence lA occurs in that position in group F6. It also appears in
other combinations in groups B2 and G-H, yielding the following paradigm.
(17)
a.
-lA
b.
-lAn
c.
-lAş
d.
-lAt
Key, p.10
3 MORPHOLOGY OF THE CAUSATIVE/INCHOATIVE ALTERNATION
Verbs formed with the simple form in (13a) are either transitive (B2) or intransitive
(F6)
transitive –lA (B2)
(18)
temizle-
‘clean’
durula-
‘rinse’
intransitive –lA (F6)
(19)
bayatla-
‘go stale’
şişmanla-
‘get fat’
-lAn and –lAş yield only intransitive verbs (G1, G2, H1)
intransitive –lAn
(20)
hastalan-
‘become sick’
G2
pislen-
‘become filthy’
H1
gerçekleş-
‘become real’
G1
kötüleş-
‘become bad’
G1
intransitive –lAş
(21)
The suffix -lAt yields only transitive verbs (F2, H1)
transitive -lAt
(22)
bayatlat-
‘cause to go stale’
F6
pislet-
‘make filthy’
H1
Key, p.11
3 MORPHOLOGY OF THE CAUSATIVE/INCHOATIVE ALTERNATION
-lA is clearly a verbalizer (little v). The other segments (-n, -ş, -t) indicate the ‘flavor’ of
little v. I therefore propose that the verbalizer (little v proper) and the flavor are
separate functional heads.
(23)
–lA
=
v
–n, –ş =
BECOME
–t
CAUSE
=
The proposed ‘flavor’ morphemes occur independently of –lA .
(24)
–n BECOME: B1, H2
ısın-
‘become hot’
tüken- ‘become exhausted’
(25)
H2
–ş BECOME: E3
sıkış- ‘become jammed’
E3
buruş- ‘become wrinkled’
(26)
–t CAUSE: F1-6, H2
ısıt-
‘heat’
H2
yıprat- ‘wear out’
The flavor morphemes never stack. Thus, sequences such as *-nt, *-şt, *-tn, *-tş
are non-occurring. (The last two are phonotactically illicit, but versions with epenthetic
vowels also do not occur: *-nIt, *-şIt). Instead, –t and –n occur in alternation, both with
-lA (H1) and with a vowel (H2).
Key, p.12
(27)
3 MORPHOLOGY OF THE CAUSATIVE/INCHOATIVE ALTERNATION
transitive
intransitive
ıs-lat-
ıs-lan-
‘get wet’
kir-let-
kir-len-
‘get dirty’
ısıt-
ısın-
‘heat’
tüket-
tüken-
‘exhaust’
yıprat-
yıpran-
‘wear out’
pis-let-
pis-len-
H1
‘get filthy’
H2
These facts suggest that, at this level, the flavor morphemes alternate on a common
base.
Hypothesis: In cases where a flavor morpheme appears to be added to the other
member of a pair which lacks a flavor morpheme, it is in fact in alternation with a null
flavor morpheme.
(28)
transitive
intransitive
a. sar-ar-t
sar-ar-ø-
√-v-CAUSE
‘make yellow’
b. giz-le-ø-
√-v-BECOME
‘become yellow’
giz-le-n-
√-v-CAUSE
√-v-BECOME
‘hide (tr.)’
‘hide (int.)’
In some cases, however, little v and flavor appear to be bundled or fused into a single
suffix.
(29)
transitive
a. boz-ø-
intransitive
boz-ul-
√-v.CAUSE
√-v.BECOME
‘ruin’
‘become ruined’
Key, p.13
b. don-dur-
3 MORPHOLOGY OF THE CAUSATIVE/INCHOATIVE ALTERNATION
don-ø-
√-v.CAUSE
√-v.BECOME
‘freeze’
‘freeze/become frozen’
Thus, little v and flavor can either be discrete or fused.
Here is a tentative catalogue of some verbal morphemes in the data set:
Isolated morphemes
(30)
little v
-lA
-Ar
-A
-I
(31)
BECOME
-n
-ş
-ø
(32)
CAUSE
-t
-ø
Fused morphemes
(33)
little v.BECOME
-Il (groups A, F3)
-ø
Key, p.14
(34)
3 MORPHOLOGY OF THE CAUSATIVE/INCHOATIVE ALTERNATION
little v.CAUSE
-DIr (groups E, G)
possibly –Ir (group C)
-ø
-lA is the Elsewhere form for pure little v. It selects for both roots (35) and whole words
(36).
root base
(35)
ıs-la-n√-v-BECOME
‘get wet’
(36)
adjective base
bayat-la-ø-
F6
stale-v-BECOME
‘go stale’
noun base
parça-la-ø-
B2
piece-v-CAUSE‘shred/shatter’
Note that –lA attaches to polysyllabic whole-word bases to form trisyllabic (or longer)
verb stems. The other little v morphemes only select for monosyllabic stems—i.e.,
roots.
Furthermore, the -n/-t alternation (H1-2) only occurs on disyllabic stems. Even
though the sequences –lAn and –lAt can occur beyond the second syllable (B2, F6), they
are never in alternation with each other anywhere but the second syllable.2 Therefore,
it seems that flavor may be a discrete morpheme only when little v is root-adjacent. At
2
In the data set, that is. Outside this set, the pair aydın-lan-/aydın-lat- ‘illuminate’ exists. It is possible that
aydın is an anomolous, disyllabic root.
Key, p.15
3 MORPHOLOGY OF THE CAUSATIVE/INCHOATIVE ALTERNATION
higher attachment sites, either little v and flavor are always fused, or they are not
distinct categories.
This would put flavor in exactly the position of the mysterious morpheme ‘M’
identified in Svenonius (2005), demarcating the boundary between the inner and outer
causative domains.
(37)
kop-ar-t-tır√-CAUS-‘M’-CAUS-
‘cause to cause to break off’
In Malagasy, ‘M’ is the prefix -f-. Nelson (to appear) has identified this with the flavor
phrase.
The puzzle mentioned above has been solved. The question was why there
wasn’t more root-adjacent allomorphy in Turkish. The answer is that, in fact, there is.
In pairs with –t in the transitive member, allomorphy is found not in the causative
morpheme, but in pure little v, which is root-adjacent.
little v allomorphy
(38)
a.
-lA (Elsewhere form)
baş-la-t-
b.
√-v-CAUSE
-Ar
F6
D
sar-ar-t√-v-CAUSE
c.
-A
uz-a-t-
F4
√-v-CAUSE
d.
-I
kur-u-t√-v-CAUSE
F5
Key, p.16
3 MORPHOLOGY OF THE CAUSATIVE/INCHOATIVE ALTERNATION
Now there is a new puzzle: If this is all correct, then in some alternations the causative
does embed an inchoative layer (E2-3, F3, G1-2), while in others it doesn’t.
Inchoative-embedding causatives
(39)
transitive
intransitive
bu-la-n-dır-
bu-la-n-
√-v-BECOME-v.CAUSE
√-v-BECOME
‘make turbid’
G2
‘become turbid’
*bulatNon-inchoative-embedding causatives
(40)
transitive
intransitive
ıs-la-t-
ıs-la-n-
√-v-CAUSE
√-v-BECOME
‘make wet’
‘become wet’
*ıslandır-
H1
We might also ask why forms such as *bulat- do not occur rather than bulandır-, or
*ıslandır- instead of ıslat-. At first glance, it looks like a straightforward case of
blocking. If ıslat- (40) is a lexical causative, it blocks the causative made with the
Elsewhere form –DIr: *ıslandır-. Where no lexical causative exists (*bulat-), Elsewhere is
free to apply: bulandır- (39).
However, on closer inspection this cannot be right. If the present analysis is
correct, the flavor morpheme –t is selecting for pure little v, while the Elsewhere
causative –DIr is selecting for a flavor phrase.
Key, p.17
3 MORPHOLOGY OF THE CAUSATIVE/INCHOATIVE ALTERNATION
(41)
CAUSE-P
vP
√P
CAUSE
-t
v
-la
DP
√
ıs-
(42)
vP
BECOME-P
v.CAUSE
-dır
vP
√P
BECOME
v
-n
-la
DP
√
bu-
In this configuration, the two morphemes are not in competition, and therefore one
cannot block the other. This is on the assumption that blocking is limited to
allomorphy at a terminal node (Embick and Marantz 2008). I make the following
proposal:
(43)
Pseudo-blocking
The existence of a functionally equivalent structure may allow a paradigm gap
to persist.
Key, p.18
3 MORPHOLOGY OF THE CAUSATIVE/INCHOATIVE ALTERNATION
In most cases, the presence or absence of an inchoative layer has little or no perceptible
semantic or syntactic effect in a causative construction. I claim that the following
representations are in most cases semantically equivalent:
(44)
CAUSE (x, [BECOME [STATE(y)]])
(45)
CAUSE (x, [STATE(y)])
If this is so, then no form is actually blocked; there is simply a paradigm gap. In cases
where there is a difference between the two representations, both structures should be
available. One type of case that makes a difference: where the inchoative is ambiguous
between a physical reading and a pyschological/affective reading.
(46)
ıs-ı-n-
H2
√-v-BECOME
(47)
a.
‘become warm (physically)’
b.
‘warm up to (emotionally)’
a.
ıs-ı-t√-v-CAUSE
‘heat (physically)’
b.
No inchoative layer
ıs-ı-n-dır-
√-v-BECOME-v.CAUSE
‘cause to warm up to (emotionally)’
Inchoative layer required
Another type of case is where the lexical causative has marked semantic drift, and an
outer causative embedding the inchoative is necessary to causativize the sense of the
intransitive verb.
(48)
a.
bay-ø√-v.CAUSE
‘bore to tears’
Key, p.19
b.
3 MORPHOLOGY OF THE CAUSATIVE/INCHOATIVE ALTERNATION
bay-ıl-
F3
√-v.BECOME
c.
‘lose consciousness, faint’
bay-ıl-t-
√-v.BECOME-CAUSE
‘cause to faint’
Note that the verb stem bayıl-, when paired with bay-, belongs in group A, but when
paired with bayılt- belongs in F3.
4
Conclusion
The decomposition of little v into a pure verbalizer (little v proper) and a flavor
head allows a fine-grained morphological analysis of the causative/inchoative
alternation in Turkish. Three derivational patterns are attested. Separate derivations
from a common base may occur at different levels: alternation of little v morphemes on
a root, or alternation of flavor morphemes on little v (among other possibilities). Pairs
that apparently support decausativization make up only 16.4% of the data set.
Also recall that in Persian, the decausativization pattern is quite rare. It is possible to
maintain decausativization only if morphological evidence is discounted (relegating
morphology to a separate module).
References
Chierchia, Gennaro, 2004. ‘A Semantics for Unaccusatives and its Syntactic
Consequences.’ In The Unaccusativity Puzzle, ed. Artemis Alexiadou, Elena
Anagnostopoulou and Martin Everaert.
Embick, David & Alec Marantz 2008. ‘Architecture and Blocking.’ Linguistic Inquiry,
Vol. 39, No. 1. Winter, pp. 1-53.
Folli, Rafaella & Heidi Harley, 2005. ‘Consuming results in Italian and English: Flavors
Key, p.20
REFERENCES
of v.’ In Aspectual inquiries, ed. by Paula Kempchinsky and Roumyana Slabakova,
95-120. Dordrecht: Springer.
Folli, Rafaella, Heidi Harley & Simin Karimi 2004. ‘Determinants of event type in
Persian complex predicates.’ Lingua vol. 115, no. 10, 1365-1401.
Göksel, Aslı and Celia Kerslake, 2005. Turkish: A Comprehensive Grammar. New York,
NY: Routledge.
Harley, Heidi, 2006. “On the causative construction.” The Handbook of Japanese
Linguistics, Shigeru Miyagawa and Mamoru Saito, ed. Oxford: OUP.
Haspelmath, Martin (1993). More on the typology of inchoative/causative verb
alternations. Causatives and Transitivity, Bernard Comrie and Maria Polinsky, ed:
87-120. John Benjamins Publishing Company, Amsterdam/Philidelphia.
Horvath, Julia, and Tal Siloni, 2009. ‘Causatives across components.’ Natural Language
& Linguistic Theory, vol. 29, no. 3, 657-704.
Jaeckel, Ralph, and Gülnur Doğanata Erciyeş (2006). A Dictionary of Turkish Verbs in
Context and by Theme. Georgetown University Press: Washington, D.C.
Kearns, Kate, 2000. Semantics. Palgrave MacMillan: New York.
Koontz-Garboden, 2009. ‘Anticausativization.’ Natural Language & Linguistic Theory.
27: 77-138.
Laks, Lior, 2011. ‘How did Hebrew ‘freeze’ (hikpi) defrost? Morphological filling of
frozen lexical entries.’ Talk given at the University of Arizona Department of
Linguistics, March 3, 2011.
Svenonius, Peter, 2005. Two domains of causatives. Talk presented at CASTL, March
10. http://www.hum.uit.no/a/svenonius/papers/Svenonius05TwoDomains.pdf
Nelson, Ryan. To appear. ‘A Principled Account of Malagasy Deverbal Nouns.’
Proceedings of the 29th West Coast Conference on Formal Linguistics.
Somerville, Mass: Cascadilla Proceedings Project
Key, p.21
APPENDIX
Appendix
A: Ø
-Il
transitive
intransitive
aç-
açıl-
‘open’
boz-
bozul-
‘ruin’
böl-
bölün-
‘divide’
bur-
burkul-
‘sprain’
dök-
dökül-
eğ-
eğil-
‘bend
göm-
gömül-
‘bury’
kır-
kırıl-
‘break’
üz-
üzül-
‘sadden’
yor-
yorul-
‘tire’
ayır-
ayrıl-
‘separate’
kavur-
kavrul-
‘roast’
B1: Ø
-n
transitive
intransitive
boşa-
boşan-
‘divorce’
yıka-
yıkan-
‘clean’
B2: Ø
-n
transitive
intransitive
gizle-
gizlen-
‘hide’
sakla-
saklan-
‘hide’
bağla-
bağlan-
‘attach’
topla-
toplan-
‘gather’
parçala-
parçalan-
‘shred/shatter’
kilitle-
kilitlen-
‘lock’
temizle-
temizlen-
‘clean’
durula-
durulan-
‘rinse’
toparla-
toparlan-
‘tidy’
yarala-
yaralan-
‘injure’
‘pour, spill’
Key, p.22
C
APPENDIX
-Ir
Ø
transitive
intransitive
artır-
art-
‘increase’
batır-
bat-
‘sink’
bitir-
bit-
‘finish’
doğur-
doğ-
‘be born’
doyur-
doy-
‘satiate’
düşür-
düş-
‘fall’
kaçır-
kaç-
‘flee’
pişir-
piş-
‘cook’
şişir-
şiş-
‘swell’
taşır-
taş-
‘overflow’
yatır-
yat-
‘lie/lay’
D:
-Ar/-Art
Ø
transitive
intransitive
çıkar-/çıkart-
çık-
‘emerge’
kopar-/kopart-
kop-
‘break off’
çöker-/çökert-
çök-
‘collapse’
göçer-/göçert-
göç-
‘cave in’
E1: -DIr
Ø
transitive
intransitive
dindir-
din-
‘subside’
doldur-
dol-
‘fill’
dondur-
don-
‘freeze’
kaldır-
kalk-
‘rise’
öldür-
öl-
‘die’
söndür-
sön-
‘extinguish’
sustur-
sus-
‘silence’
döndür-
dön-
‘turn’
durdur-
dur-
‘stop’
oldur-
ol-
‘become’
saptır-
sap-
‘turn’
Key, p.23
APPENDIX
E2: -DIr
Ø
transitive
intransitive
sevindir-
sevin-
‘cheer up’
uyandır-
uyan-
‘awaken’
E3: -DIr
Ø
transitive
intransitive
değiştir-
değiş-
‘change’
dönüştür-
dönüş-
‘metamorphose’
oluştur-
oluş-
‘form’
sıkıştır-
sıkış-
‘jam’
buruştur-
buruş-
‘crumple’
barıştır-
barış-
E4: -DIr
Ø
transitive
intransitive
acıktır-
acık-
‘get hungry’
biriktir-
birik-
‘accumulate’
F1: -t
Ø
transitive
intransitive
gebert-
geber-
‘croak (die)’
karart-
karar-
‘blacken’
kızart-
kızar-
‘redden/fry’
sarart-
sarar-
‘yellow’
yaşart-
yaşar-
‘water (eyes)’
F2: -t
Ø
transitive
intransitive
çıldırt-
çıldır-
‘go crazy’
delirt-
delir-
‘go crazy’
belirt-
belir-
‘manifest’
‘make peace’
Key, p.24
APPENDIX
F3: -t
Ø
transitive
intransitive
alçalt-
alçal-
‘lower’
azalt-
azal-
‘decrease’
boşalt-
boşal-
‘empty’
çoğalt-
çoğal-
‘increase’
düzelt-
düzel-
‘straighten’
kısalt-
kısal-
‘shorten’
yükselt-
yüksel-
‘rise’
F3: -t
Ø
transitive
intransitive
ayılt-
ayıl-
‘come to’
bayılt-
bayıl-
‘faint’
darılt-
darıl-
‘offend’
eksilt-
eksil-
‘reduce’
yanılt-
yanıl-
‘go astray’
F4: -t
Ø
transitive
intransitive
susat-
susa-
‘thirst’
kaynat-
kayna-
‘boil’
uzat-
uza-
‘extend’
F5: -t
Ø
transitive
intransitive
büyüt-
büyü-
‘grow’
erit-
eri-
‘melt’
eskit-
eski-
‘age’
kurut-
kuru-
‘dry’
soğut-
soğu-
‘chill’
Key, p.25
F6: -t
APPENDIX
Ø
transitive
intransitive
bayatlat-
bayatla-
‘go stale’
şişmanlat-
şişmanla-
‘fatten’
gerilet-
gerile-
‘regress’
ihtiyarlat-
ihtiyarla-
‘age’
ucuzlat-
ucuzla-
‘get cheap’
yavaşlat-
yavaşla-
‘slow’
zayıflat-
zayıfla-
‘get thin’
başlat-
başla-
‘start’
G1: -DIr
Ø
transitive
intransitive
buharlaştır-
buharlaş-
‘vaporize’
yerleştir-
yerleş-
‘settle’
zorlaştır-
zorlaş-
‘get difficult’
yaklaştır-
yaklaş-
‘approach’
birleştir-
birleş-
‘unite’
fenalaştır-
fenalaş-
gerçekleştir-
gerçekleş-
‘realize’
iyileştir-
iyileş-
‘improve’
kolaylaştır-
kolaylaş-
‘facilitate’
kötüleştir-
kötüleş-
‘deteriorate’
uzaklaştır-
uzaklaş-
‘distance’
‘deteriorate’
Key, p.26
G2: -DIr
APPENDIX
Ø
transitive
intransitive
bulandır-
bulan-
‘become turbid’
ayaklandır-
ayaklan-
‘revolt’
evlendir-
evlen-
‘marry’
hastalandır-
hastalan-
‘get sick’
umutlandır-
umutlan-
‘get hopeful’
yaşlandır-
yaşlan-
‘age’
neşelendir-
neşelen-
‘cheer up’
onurlandır-
onurlan-
‘honor’
öfkelendir-
öfkelen-
‘enrage’
heyecanlandır-
heyecanlan-
‘excite’
hüzünlendir-
hüzünlen-
‘sadden’
kaygılandır-
kaygılan-
‘get anxious’
H1: -t
-n
transitive
intransitive
ıslat-
ıslan-
‘get wet’
kirlet-
kirlen-
‘get dirty’
pislet-
pislen-
‘get filthy’
H2: -t
-n
transitive
intransitive
ısıt-
ısın-
‘heat’
tüket-
tüken-
‘exhaust’
yıprat-
yıpran-
‘wear out’
kapa-/kapat-
kapan-
‘open’
I:
unique patterns
transitive
intransitive
yak-
yan-
‘burn’
dağıt-
dağıl-
‘scatter’
kurtar-
kurtul-
‘get free’
Key, p.27
J
et-
APPENDIX
ol- (light verb alternation)
transitive
intransitive
hall-et-
hall-ol-
‘solve’
icat et-
icat ol-
‘invent’
ihraç et-
ihraç ol-
‘export’
israf et-
israf ol-
‘waste’
kabul et-
kabul ol-
‘accept’
kahr-et-
kahr-ol-
‘damn’
kayb-et-
kayb-ol-
‘lose’
kayd-et-
kayd-ol-
‘record’
meşgul et-
meşgul ol-
‘occupy’
rahatsız et-
rahatsız ol-
‘bother’
sünnet et-
sünnet ol-
‘circumcise’
tahrik et-
tahrik ol-
‘arouse’
tamir et-
tamir ol-
‘repair’
tedavi et-
tedavi ol-
‘treat’
teslim et-
teslim ol-
‘surrender’
var et-
var ol-
‘come into existence’
yok et-
yok ol-
‘cease existence’
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Related manuals

Download PDF

advertisement

Languages