Another look at velar deletion in Turkish, with special ... - Linguistics
Another look at velar deletion in Turkish, with special ... - Linguistics
Another look at velar deletion in Turkish, with special ... - Linguistics
You also want an ePaper? Increase the reach of your titles
YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.
16<br />
F<strong>in</strong>nish assibil<strong>at</strong>ion, while exemplified by many words, is actually highly<br />
morphologically constra<strong>in</strong>ed. Both Andrew Dolbey and Paul Kiparsky have observed (pc,<br />
1998) th<strong>at</strong> the assibil<strong>at</strong><strong>in</strong>g /t/ consonant cited <strong>in</strong> most examples is actually a s<strong>in</strong>gle, very<br />
productive morphological form<strong>at</strong>ive used <strong>in</strong> denom<strong>in</strong>al and onom<strong>at</strong>opoetic verb deriv<strong>at</strong>ion<br />
and <strong>in</strong> the adapt<strong>at</strong>ion of loan words. Examples provided to us by Dolbey <strong>in</strong>clude [[tila-T]-i]<br />
tilasi ‘ordered’, [[töppä-T]-i] töppäsi ‘behaved stupidly’, and [[digga-T]-i] diggasi ‘thought<br />
was really cool’ (see also Inkelas 2000). Anttila (2006), cit<strong>in</strong>g work published <strong>in</strong> F<strong>in</strong>nish by<br />
Fred Karlsson, notes th<strong>at</strong> Assibil<strong>at</strong>ion is triggered by three suffixes: “plural /-i/, e.g.<br />
/vuote-i-nA/ → vuos<strong>in</strong>a ‘year-PL-ESS’, past tense /-i/, e.g. /huuta-i-vAt-kO/ → huusiv<strong>at</strong>ko<br />
‘shout-PAST-3P.PL-QUE’, and superl<strong>at</strong>ive /-impA/, e.g. /uute-impA-nA/ → uusimpana<br />
‘new-SUP-ESS’” (p. 900). But Assibil<strong>at</strong>ion does not apply generally before all i-<strong>in</strong>itial<br />
suffixes, e.g. -iist and -iiv. For example, Kiparsky (2003) cites the forms vok<strong>at</strong>-iivi-lla<br />
‘voc<strong>at</strong>ive’ and nom<strong>in</strong><strong>at</strong>-iiv-lla (p. 116), <strong>in</strong> which a morphologically derived /t-i/ sequence<br />
does not undergo Assibil<strong>at</strong>ion. To this list of exceptions Anttila (2006) adds several others:<br />
“the deriv<strong>at</strong>ional suffix /-ime/ ‘<strong>in</strong>strument’, e.g. /lentä-ime-n/ → lentimen / *lensimen ‘fly<strong>in</strong>strument-GEN’<br />
, and the sign<strong>at</strong>ure /-isi/ ‘conditional’, e.g. /tunte-isi/ → tuntisi (*tunsisi)<br />
‘feel-COND’.” Anttila also notes th<strong>at</strong> “the deriv<strong>at</strong>ional suffix /-<strong>in</strong>en/ only triggers<br />
Assibil<strong>at</strong>ion optionally, e.g. /vete-<strong>in</strong>en/ → ves<strong>in</strong>en ∼ vet<strong>in</strong>en ‘w<strong>at</strong>ery’” (p. 901). Anttila<br />
concludes th<strong>at</strong> Assibil<strong>at</strong>ion is a stem-level process. The broader po<strong>in</strong>t, however, is th<strong>at</strong> even<br />
<strong>in</strong> this parade example of a morphologically derived environment condition, it is not<br />
enough to say th<strong>at</strong> the environment must be morphologically complex. Once a sufficiently<br />
precise list<strong>in</strong>g of morphological contexts has to be provided, the overarch<strong>in</strong>g derived<br />
environment generaliz<strong>at</strong>ion becomes extraneous <strong>at</strong> best.<br />
Phonologically derived environment condition<strong>in</strong>g is equally hard to p<strong>in</strong> down. In<br />
F<strong>in</strong>nish, as noted earlier, word-f<strong>in</strong>al rais<strong>in</strong>g of /e/ to [i], a phonological altern<strong>at</strong>ion, cre<strong>at</strong>es a<br />
derived environment for Assibil<strong>at</strong>ion: /vete/ ‘w<strong>at</strong>er’ → |veti| → [vesi]. But it is not the case<br />
th<strong>at</strong> just any phonologically derived /ti/ sequence is subject to Assibil<strong>at</strong>ion. Consonant<br />
grad<strong>at</strong>ion, which degem<strong>in</strong><strong>at</strong>es <strong>in</strong>ter-sonorant voiceless plosives <strong>in</strong> closed syllables, can<br />
convert /…tti…/ str<strong>in</strong>gs to /…ti…/, meet<strong>in</strong>g the structural description for Assibil<strong>at</strong>ion.<br />
However, examples like ott-i-n → ot<strong>in</strong> (*os<strong>in</strong>) ‘take-PAST’ show th<strong>at</strong> Assibil<strong>at</strong>ion is not<br />
triggered <strong>in</strong> this environment (Anttila 2006:896).<br />
Miller (1975) raises the <strong>in</strong>terest<strong>in</strong>g example of West Greenlandic, which has an<br />
assibil<strong>at</strong>ion altern<strong>at</strong>ion th<strong>at</strong> on first <strong>in</strong>spection presents a better case of a derived<br />
environment rule than does <strong>Turkish</strong> <strong>velar</strong> <strong>deletion</strong>. In West Greenlandic, /t/ assibil<strong>at</strong>es to /s/<br />
<strong>in</strong>tervocalically after /i/ <strong>in</strong> morphologically derived environments, both stem-f<strong>in</strong>ally (22a)<br />
and suffix-<strong>in</strong>itially (22b). While epenthetic [i] can cre<strong>at</strong>e the <strong>in</strong>tervocalic environment<br />
required for assibi<strong>at</strong>ion to apply (22c), an epenthetic [i] cannot serve as the immedi<strong>at</strong>e,<br />
preconsonantal trigger for epenthesis (22d):<br />
(22)<br />
a. /tikit-aq/ tikisaq ‘come-pass.ppl’<br />
cf. /tikit-li/ tikitli ‘come-3sg.opt’<br />
cf. /titit-tuq/ tikittuq ‘come-act.ppl’<br />
b. /iki-tit/ ikisit ‘your (sg.) wounds<br />
cf. /iga-tit/ ig<strong>at</strong>it ‘your (sg.) cook<strong>in</strong>g pots’