The feature [nasal] is surely spoken Karthik Durvasula University of Delaware Introduction

¾ Despite having acoustic/perceptual correlates and diachronic effects similar to nasals, aspirates do not synchronically behave like nasals.

Future Direction

¾ Some recent studies on phonological features have questioned the traditional generative phonology claim that features are innate.

¾ This observation leads to:

¾ Mielke (2004): • Traditional natural classes account for at most 75% of the observed cross-linguistic phonological phenomena. • Emergentist feature theories, as per which, features are just emergent abstractions that are influenced by a variety of phonetic, cognitive and social factors can lead to a better understanding of cross-linguistic patterns.

Result A: Nasal harmony and long-distance nasal consonant agreement are triggered only by segments which are articulatorily nasal.

¾ What about [strident]? • Usually seen to have an acoustic/perceptual target; No simple articulatory correlate; • Is motor-bootstrapping sufficient/too constrained? • Are innate features the only way?

¾ Lin (2005): • similar conclusion based on modern computational and statistical techniques. • broad phonetic (acoustic) classes can be learnt through unsupervised learning algorithms. ¾ This paper shows that: • unconstrained emergentist approaches are incapable of accounting for the bias observed in phonological phenomena. • Specifically: nasal harmony and long distance nasal consonant harmony patterns show a strong articulatory bias – aspirated segments despite having acoustic/perceptual effects similar to typical nasal segments never trigger nasal harmony.

The Phonetics ¾ Close phonetic connection between aspirates and nasals. Aspiration and Nasality have similar Acoustic Correlates/Effects ¾ Aspiration and nasalization have similar acoustic correlates/effects on vowels ((Ohala & Amador (1981), Ohala (1983), Ohala & Ohala (1993)): 1. (a) Decreased first formant (F1) amplitudes. (b) Increased first formant (F1) bandwidth. ¾ Arai (2006) advanced the following schematic to capture the acoustic/perceptual overlap.

etc.

Low-freq. peak

Phonological Adjacency Effects ¾ Aspirate segments have phonological effects on adjacent segments that are similar to those of nasals. Rhinoglottophilia or Spontaneous Nasalisation ¾ Like nasals, aspirates (including high-airflow segments like fricatives and affricates) have (diachronically) been observed to induce nasality next to themselves (2) (Matisoff (1975), Ohala & Ohala (1993)). 2. Rhinoglottophilia - Modern Hindi (from Ohala and Ohala (1993)) Sanskrit Old Hindi Modern Hindi (a) pakʂa pa:kh pəkha ‘a side’ u:ča ‘high’ (b) uččakaN/A Nasals Induce Aspiration in Swahili ¾ Also, observed cases of aspiration emerging on (voiceless) consonants adjacent to nasal consonants in Swahili (3) (Botma 2004). 3. Aspiration in Swahili (Botma 2004) (a) /N+pepo/ phepo ‘spirits’ (c) /N+buzi/ mbuzi ‘goat’ (d) /N+dege/ ndege ‘bird’ (b) /N+tembo/ thembo ‘elephant’

¾ Sundanese: • Nasal harmony blockers include segments phonetically underspecified for nasality (5) (Cohn 1993a). • /l/ blocks nasality (5b), but is variably nasalized depending on the phonetic context. • The harmony blockers thus have no obvious common phonetic property. Harmony must be an abstract process. 5. Sundanese Nasal Harmony (Cohn (1993a)) (a) ɲiar → ɲiar ‘seek (sing.)’ (b) ŋuliat → ŋuliat ‘stretch (active)’ (c) ɲ=al=iar → ɲa l iar ‘seek (pl.)’ Result B: Nasal harmony must be an abstract, i.e., phonological, phenomenon. Where is the bias? ¾ Results (A) and (B) force us to conclude that: • The representational system - i.e., the feature system - makes reference to articulatory similarities as opposed to acoustic/perceptual similarities with respect to nasal harmony. Result C: the phonological feature [nasal] is (at least) defined articulatorily.

Wide B1

High-freq. noise

Fig. 1. Acoustic/Perceptual cues related to nasality and breathiness in vowels.

Perceptual Effects ¾ Vowel portions immediately adjacent to aspirate segments are perceived to be nasal. ¾ Ohala & Amador (1981): listeners judged vowel stimuli made by iterating single pitch periods of portions of vowels immediately next to voiceless fricatives (high-airflow segments produced with greaterthan-normal glottal opening) to be about as nasalized as comparable periods made from vowel margins near nasal consonants even though the former show “no physiological nasalization” (Ohala 1983).

¾ Towards a better understanding of phonological features. • Are span/harmony processes better hunting grounds for phonological features than immediate-adjacency-based processes which are more susceptible to various peripheral effects such as co-articulatory or misperception effects? • Rethink many proposed features.

4. Harmony in Guaraní (a) tupá → tupá ‘god’ (Rivas 1975) (b) popí → popí ‘to peel, strip’ (Walker 1998)

Breathiness

Nasality Pole-zero pairs

Fig. 1. Listeners' judgements of degree of nasality of iterated vowel (left bar of each pair from middle of vowel; right bar from the end); Ohala 1983)

Nasal harmony is clearly Phonological not Phonetic ¾ Guaraní: • Nasal harmony ‘skips’ simple voiceless-stop segments (Rivas (1975), Walker (1998)); • The harmony rule cannot be the phonetic spreading of nasality to adjacent segments (4). • The process here must be at a more abstract phonological level.

Observed Bias in the Phonology Articulatory bias in nasal harmony and long-distance nasal consonant agreement ¾ There are NO observed cases of aspirated segments partaking in long-distance nasal consonant agreement or nasal harmony. ¾ Potentially problematic • Glottal fricatives in Aguaruna and Arabela spread nasality along with the nasal segments. • But, Trigo (1988), Walker & Pullum (1999): These segments are underlyingly nasal; thus, not counterexamples .

Conclusion

References Arai, T. (2006). Cue parsing between nasality and breathiness in speech perception. Acoust. Sci. & Tech. 27, 5. Botma, B. (2004). Phonological Aspects of Nasality: an Element-based Dependency Account. Doctoral dissertation, University of Amsterdam. Cohn, Ab. (1993a). “The Status of Nasalized Continuants”, in Marie K. Huffman and Rena A. Krakow eds., Phonetics and Phonology – Nasals, Nasalization, and the Velum, Academic Press, 329-368. Halle, M. and Stevens, K.N. (1991). “Knowledge of Language and the Sounds of Speech”, in Sundberg, J., Nord, L., and Carlson, R. eds., Music, Language, Speech and Brain, MacMillian Press, London, 1-19. Lin, Y. (2005). Learning Features and Segments from Waveforms: A Statistical Model of Early Phonological Acquisition. UCLA Dissertation. Lindblom, B. (1999). “Emergent Phonology”, manuscript. Matisoff, J. A. (1975). "Rhinoglottophilia: The Mysterious Connection between Nasality and Glottality", in C.A. Ferguson, L.M. Hyman and J.J. Ohala eds., Nasálfest: Papers from a Symposium on Nasals and Nasalization, Universals Language Project, Stanford University, Stanford, 265-287. Mielke, J. (2004). “Moving beyond innate features: a unified account of natural and unnatural classes”, NELS 35. Ohala, J.J. (1983). The Phonological ends justifies the means. In S. Hattori & K. Inoue (eds.), Proc. of the XIIIth Int. Cong. of Linguists, Tokyo, 29 Aug. - 4 Sept. 1982. Tokyo. [Distributed by Sanseido Shoten.] 232 - 243. Ohala, J.J., & Amador, M., (1981). “Spontaneous Nasalization”, Journal of the Acoustical Society of America, 68, S54-S55. Ohala, J.J., & Ohala, M. (1993). “The Phonetics of Nasal Phonology: Theorems and Data”, in Marie K. Huffman and Rena A. Krakow eds., Phonetics and Phonology – Nasals, Nasalization, and the Velum, Academic Press, 225-250. Rivas, A. M., (1975). “Nasalization in Guarani”, NELS 6. Trigo, R. L. (1988). On the phonological derivation and behavior of nasal glides. Cambridge, Ma: MIT Dissertation. Walker, R. (1998). Nasalization, Neutral Segments, and Opacity Effects. PhD dissertation, University of California, Santa Cruz. [Published by Garland, New York, 2000]. Walker, R. & Pullum G. K. (1999). Possible and Impossible Segments. Language 75 764780.

¾ Reason to believe that there is a featural cognitive bias: [nasal] is (at the very least) defined with an articulatory target. ¾ Unconstrained Emergentist feature theories: • cannot capture the articulatory bias observed in nasal harmony and long-distance nasal consonant agreement. • not clear why acoustically similar effects of glottal aspirates are not over-generalized, beyond immediate adjacency effects, to behave like regular harmony. ¾ However, the present conclusion readily supports: • Innatist articulatorily features (Halle and Stevens 1991). • Or, emergent features guided by ‘motor boot-strapping’ (Lindblom 1999).

Acknowledgements I would like to thank Dr. William Idsardi, Dr. Jeffrey Heinz, and Timothy McKinnon for insightful discussions.

Author’s Contact Information [email protected]