Markedness and Complex Stops: Evidence from Simplification Processes 1 Nick Danis Rutgers University nick.danis@rutgers.edu WOCAL 8 Kyoto, Japan August 21-24, 2015 1 Introduction (1) Complex segments: segments with two unordered phonological place features (Chomsky and Halle 1968; Clements and Hume 1995; Sagey 1986, a.o.) Labial-velars: kp, gb Labial-coronals: tp, db Clicks:!,, (2) Simplification Process A process where a complex segment in the input is realized as a simple segment in the output Attested simplification processes: Labial-velars: /KP 2 / [P] Amele, Efik, Ibibio Labial-coronals: /TP/ [T] Margi Clicks: /!/ [K] Fwe, Yeyi (3) Generalization: The choice of place to which a doubly-articulated stop reduces is determined by universal markedness in the grammar: a segment reduces to its least marked but still faithful place (4) Universal Markedness Hierarchy dorsal > labial > coronal > glottal (de Lacy 2006; Lombardi 2001; Prince and Smolensky 1993/2004) (5) Clicks are subject to an additional faithfulness constraint: reduction to more marked [dorsal] place is possible (6) Representations are simplified: no abstract primary vs. secondary place distinction is necessary 1 Thanks to Akin Akinlabi, Alan Prince, Bruce Tesar, Paul de Lacy, Mike Cahill, John Roberts, Bonny Sands, Hilde Gunnink, Will Bennett, Hope McManus, Natalie DelBusso, and Paula Houghton for all help, comments, and pointers along the way. All mistakes are my own. Previous versions of this talk were given at ACAL46 and at the Workshop on Formal Typologies at Rutgers University. 2 Abbreviations throughout: capital P, T, and K indicate a labial, coronal, or dorsal stop respectively regardless of voicing. KP, TP, and other combinations of these symbols indicate a complex segment, never a cluster. The symbol for the alveolar click! represents any coronal-dorsal click and not specifically one with an alveolar anterior articulation.
Markedness and Complex Stops 2 2 Representations/GEN S (7) The relevant features for the segments in question are given in (8). (8) Features Place [voice] [lingual] P [labial] +/ T [coronal] +/ K [dorsal] +/ TP [labial], [coronal] +/ KP [labial], [dorsal] +/! [coronal], [dorsal] +/ + (9) No specific geometry is assumed: Segments are sets of features Work of class nodes is done by constraint definitions (Padgett 1995a, b, 2002) (10) Specifically in complex segments, no notion of "primary place" or "major articulator" is encoded in the representation (cf. Anderson 1976; Sagey 1986) (11) Example representation: [kp] (12) [lingual] feature: [+lingual] segments are produced with the velaric/lingual airstream mechanisms (clicks) (Halle 1995; Miller 2007, 2011) 3 Constraints/CON S 3.1 Markedness [dorsal] [-voice] [labial] [-lingual] (13) Markedness constraints are based on a universal hierarchy (de Lacy 2006; Lombardi 2001; Prince and Smolensky 1993/2004): [dorsal] > [labial] > [coronal] where > signifies more marked than (14) Constraints are defined stringently rather than in a fixed universal ranking (de Lacy 2006): all three constraints assigns a violation to dorsals, only two assign a violation to labials, and only one assigns a violation to coronals.
Markedness and Complex Stops 3 (15) [glottal] is ignored on the scale her for the sake of simplicity (16) m:kpt "Don't be dorsal, labial, or coronal" Assign one violation: for every segment S where [dor] S and for every segment S where [lab] S and for every segment S where [cor] S (17) m:kp "Don't be dorsal or labial" Assign one violation: for every segment S where [dor] S and for every segment S where [lab] S (18) m:k "Don't be dorsal" Assign one violation: for every segment S where [dor] S (19) Markedness Violation Tableau m.kpt m.kp m.k T 1 0 0 P 1 1 0 K 1 1 1 TP 2 1 0 KP 2 2 1! 2 1 1 (20) m.kpt only differentiates complex segments (2 violations) from simple segments (1 violations): it is a derived *ComplexSegment constraint 3.2 Faithfulness (21) Place is subject to a general place faithfulness constraint. (22) f:place "Don't add or remove place features" If S in is a segment in the input, and S out is a segment in the output, and S in and S out are in correspondence Assign one violation for every x such that: x S in & x S out and Assign one violation for every x such that: x S out & x S in where x {[lab], [cor], [dor]}
Markedness and Complex Stops 4 (23) Violation Profile for f:place In Out f:place Comment kp t 3 [dor] and [lab] not in output, [cor] not in input kp p 1 [dor] not in output kp k 1 [lab] not in output kp kp 0 No change in place (24) Ident (McCarthy and Prince 1995) constraints for features [lingual] and [voice]: (25) f:[lingual] "Don't change the value for [lingual]" If S1 is a segment in the input, and S2 is a segment in the output, and S1 and S2 are in correspondence Assign one violation if S1 is [αlingual] and S2 is [ αlingual] (26) f:[voice] "Don't change the value for [voice]" If S1 is a segment in the input, and S2 is a segment in the output, and S1 and S2 are in correspondence Assign one violation if S1 is [αvoice] and S2 is [ αvoice] (27) Positional faithfulness is also necessary: (Beckman 1998) f-ons:place f-ons:[lingual] f-ons:[voice] 4 Simplification Processes (28) Simplification processes involving complex segments: Language Alternation Process Least marked? Source Amele gb~p Reduction Yes Roberts (1987) Efik kp~p Reduction Yes Welmers (1973) Ibibio kp~p Reduction Yes Kaufman (1968) Aghem gb~b, kp~p (rare) Agreement? Yes Hyman (1979) Margi tp~t Reduction Yes Hoffman (1963) Fwe!~K Reduction No Gunnink (in press) Yeyi!~K Reduction No Seidel (2008) 4.1 Labial-velars (29) Amele (Papua New Guinea) /gb/ is realized as [p] in coda/final position (Roberts 1987), see also (Cahill 1999, 2000)
Markedness and Complex Stops 5 (30) In final position: (Roberts 1987: 346) /d/ [t] /b/, /gb/ [p] /g/ [k] (31) Morphosyntactic alternation, based on position of TAM marker: (32) Realization as complex segment: a. /h+ogb+ona/ [ho'.gbɔ.nǝ] 'we are coming' (present) b. /f+ogb+ona/ [fo'.gbɔ.nǝ] 'we are seeing' (present) (33) Simplification to [p] in final position: a. /h+ol+ogb/ [hɔ'.lɔp] we used to come' (past habitual) b. /f+ol+ogb/ [fɔ'.lɔp] we used to see' (past habitual) (34) Efik and Ibibio (Cross River, Nigeria) Labial-velar [kp] and simple labial [p] are in complementary distribution: Welmers (1973), see also Cahill (2000) (35) Phonotactic restriction: "[p] after a pause or juncture [syllable boundary] is a coarticulated bilabial velar stop, voiceless, unaspirated [kp] before a pause it is a bilabial stop, unaspirated, or in careful speech weakly aspirated, voiceless, and in this position like t, k often unreleased [p]." (Ibibio, Kaufman 1968: 44) (36) Section Summary: labial-velars reduce to simple labials, never to simple dorsals (see also Cahill 2000, 2006) 4.2 Labial-coronals (37) In Margi (Chadic, Nigeria), speakers "reduce initial compound consonants to simple consonants, especially in a more colloquial type of speech" Hoffman (1963: 43), see also Sagey (1986) (38) Simplification of labial-coronals: a. /bdəәli/ [dəәli] 'Dille' b. /ptəәl/ [təәl] 'chief' (39) Labial coronals reduce to simple coronals, never to simple labials (nb: Margi only known case)
Markedness and Complex Stops 6 5 Rankings 5.1 Why simplify? (40) In order for a language to allow complex stops at all, some place faithfulness constraint must dominate m.kpt, the derived *ComplexSegment constraint. (41) General ranking for simple languages (e.g. English) m:kpt f:place m:kp m:k (42) General ranking for complex languages f:place m:kpt m:kp m:k (43) For Amele/Efik/Ibibio-type languages, complex segments are allowed only in onsets: f-onset:place m:kpt f:place m:kp. m:k (44) For Margi, Fwe, and Yeyi, the alternation is a type of intra-speaker free-variation. Point of variation could be variable ranking between f:place and m:kpt. 5.2 Restricted Typology (45) /KP/ [K] is Harmonically Bounded /kp/ m.kpt m.kp m.k f.place m.kpt/[+voi] f:[voi] k 1 1 1 1 0 0 p 1 1 0 1 0 0 t 1 0 0 3 0 0 kp 2 2 1 0 0 0 (46) /TP/ [P] is Harmonically Bounded /tp/ m.kpt m.kp m.k f.place m.kpt/[+voi] f:[voi] p 1 1 0 1 0 0 t 1 0 0 1 0 0 tp 2 1 0 0 0 0
Markedness and Complex Stops 7 6 Clicks 6.1 Simplification Processes (47) Fwe (Namibia, Zambia, Bantu K402) Clicks are in free variation with their simple dorsal counterparts (Gunnink in press) (48) Fwe alternations (Gunnink in press: (14, 16, 18)): [ ] [k] [kùǀàpùrà] ~ [kùkàpùrà] 'to tear' [g ] [g] [mùɡǀênè] ~ [mùɡênè] 'thin (person)' [ŋ ] [ŋ] [nɡ ɔŕɛ zà] ~ [ŋɔŕɛ zà] 'resin' (49) Voice and nasality features remain; process is loss of a single place feature: /ŋ / [ŋ] [dorsal] [+nasal] [dorsal] [+nasal] [+voice] [+lingual] [+voice] [+lingual?] [coronal] (50) Cannot only be change in value of [lingual]: this predicts a coronal-dorsal doublyarticulated stop (if admitted in GEN) (51) Seidel (2008) reports a similar alternation for the related language Yeyi. 6.2 Click Faithfulness (52) Reduction in clicks does not fit the generalization of simplification to the least marked place. (53) Two hypotheses: Clicks are subject to a different markedness scale (rejected) Clicks are subject to additional faithfulness constraint(s) (accepted) (54) A conflicting markedness scale based on [+lingual] segments cannot explain these reduction processes, because markedness by definition is subject only to outputs, and the resulting outputs are [-lingual] (i.e. non-clicks) (55) Instead, an additional faithfulness constraint is assumed that takes advantage of the fact that clicks are featurally distinct from doubly-articulated stops
Markedness and Complex Stops 8 (56) f:k/[lingual] "Preserve [dorsal] if input segment is click" If S1 and S2 are corresponding input/output segments and S1 is [+lingual], Assign 1 violation if [dorsal] S1 and [dorsal] S2. (57) Conceptually, this constraint is based on the relationship between the [+lingual] feature requiring the velaric/lingual airstream mechanism (and thus a dorsal closure) (Halle 1995; Miller 2007, 2011) (58) Clicks are now predicted to reduce either to [dorsal] or to the lesser marked place, depending on ranking 7 Summary (59) Doubly-articulated stops reduce to the less marked but still faithful place: KP P TP T (60) Clicks can reduce to the more-marked dorsal place due to their [+lingual] specification and an additional faithfulness constraint:! K (61) The reduction of doubly-articulated stops is accomplished with independently-needed mechanisms in the grammar (universal markedness) (62) Constraint definitions take the place of feature class nodes and enriched representations, following Feature Class Theory (Padgett 1995a, b, 2002) (63) A constraint banning complex segments (m:kpt) is derived from a general constraint building mechanism (stringent markedness constraints) rather than simply assumed References Anderson, Stephen R. 1976. On the description of multiply-articulated consonants. Journal of Phonetics, 4, 17-27. Beckman, Jill. 1998. Positional Faithfulness. University of Massachusetts Amherst. Cahill, Michael. 1999. Aspects of the phonology of labial-velar stops. Studies in African Linguistics, 28(2), 155-184. Cahill, Michael. 2000. Positional Contrast and Labial-Velars. OSU Working Papers in Linguistics. Cahill, Michael. 2006. The Place of Labial-Velars. Paper presented at Annual Conference on African Linguistics, Eugene, OR. April 7-9. 2006. Chomsky, Noam, & Morris Halle. 1968. The Sound Pattern of English. Cambridge, Massachusetts: The MIT Press.
Markedness and Complex Stops 9 Clements, G. N., & Elizabeth V. Hume. 1995. The Internal Organization of Speech Sounds. In John Goldsmith, Jason Riggle & Alan C. L. Yu (Eds.), The Handbook of Phonological Theory: Blackwell Publishing. de Lacy, Paul. 2006. Markedness: Reduction and Preservation in Phonology. Cambridge, UK: Cambridge University Press. Gunnink, Hilde. in press. Click loss and click insertion in Fwe. Halle, Morris. 1995. Feature Geometry and Feature Spreading. Linguistic Inquiry, 26(1), 1-46. Hoffman, Carl. 1963. A Grammar of the Margi Language. London: Oxford University Press. Hyman, L.M. 1979. Part 1: Phonology and Sound Structure. In L.M. Hyman (Ed.), Aghem Grammatical Stricture. Los Angeles: University of Southern California. Kaufman, Elaine Marlowe. 1968. Ibibio Grammar. (Ph.D.), University of California, Berkeley. Lombardi, Linda. 2001. Why Place and Voice are different: Constraint-specific alternations in Optimality Theory. Segmental phonology in Optimality Theory: Constraints and Representations. McCarthy, John, & Alan Prince. 1995. Faithfulness and Reduplicative Identity. Ms. Miller, Amanda. 2007. The Phonology of Click Consonants. Ms., Cornell University. Miller, Amanda. 2011. The Representation of Clicks. In Marc van Oostendorp, Colin J. Ewen, Elizabeth Hume & Keren Rice (Eds.), The Blackwell Companion to Phonology (pp. 416-439): John Wiley & Sons. Padgett, Jaye. 1995a. Feature Classes. In Jill Beckman, S. Urbanczyk & L. Walsh (Eds.), Papers in Optimality Theory. Amherst, MA: University of Massachusetts. Padgett, Jaye. 1995b. Partial Class Behavior and Nasal Place Assimilation. Proceedings of the Arizona Phonology Conference: Workshop on Features in Optimality Theory. Tucson, AZ: University of Arizona Department of Linguistics. Padgett, Jaye. 2002. Feature Classes in Phonology. Language, 78(1), 81-110. Prince, Alan, & Paul Smolensky. 1993/2004. Optimality Theory: Constraint Interaction in Generative Grammar. Malden, MA, & Oxford: Blackwell Publishing. Roberts, John R. 1987. Amele. Sydney: Croom Helm. Sagey, Elizabeth. 1986. The Representation of Features and Relations in Non-Linear Phonology. (PhD), Massachusetts Institute of Technology. Seidel, Frank. 2008. A Grammar of Yeyi. Köln: Rüdiger Köppe Verlag. Welmers, Wm. E. 1973. African Language Structures: University of California Press.