sign in sign up
<<
Home , Cognitive linguistics, Reduplication in Russian

ICLC-14 Book of Abstracts Theme sessions

PHONOLOGY IN COGNITIVE

Geoffrey S. Nathan & Jose A. Mompean* Wayne State University, University of Murcia* [email protected], [email protected]*

Phonology is in no way less conceptual than other areas of linguistic reseach such as or . Yet it has been much neglected in cognitive linguistics, although a few researchers have explored various ways of rethinking phonology within cognitive linguistics (see e.g. Mompean, 2014; Mompean & Mompean-Guillamón, 2012; Nathan, 2008, 2015; Nesset, 2008; Taylor, 2002). In order to foster discussion and work on phonology within the cognitive linguistic framework, the proposed theme session aims to bring together contributions dealing with the role of phonology in specific or (spoken/sign) in general as well as phonological issues from a cognitive/functional linguistic perspective. These include, among others, and prototype effects, the phonology of symbolic units and constructions, usage-based aspects of phonology, phonetic, sociolinguistic and cultural motivation of phonological units, in phonology, relationships between phonology and other gestural and non-verbal communication, etc.

References

Mompean, J.A. (2014). Cognitive Linguistics and Phonology. In J.R. Taylor & J. Littlemore (eds), The Bloomsbury Companion to Cognitive Linguistics (pp. 253–276). London: Bloomsbury Publishing. Mompean, J. A., & Mompean-Guillamón, P. (2012). La fonología cognitiva. In I. Ibarretxe-Antuñano & J. Valenzuela (eds.), Lingüística Cognitiva (pp. 307–326). Barcelona: Anthropos. Nathan, G.S. (2008). Phonology: A Introduction. Amsterdam: John Benjamins. Nathan, G. S. (2015). Phonology. In E. Dąbrowska & D. Divjak (eds), Handbook of Cognitive Linguistics (pp. 253–273). Berlin & New York: Mouton de Gruyter. Nesset, T. (2008). Abstract Phonology in a Concrete Model. Cognitive Linguistics and the - Phonology Interface (Series Cognitive Linguistics Research [CLR]). Berlin and New York: De Gruyter Mouton de Gruyter. Taylor, J.R. (2002). Phonological Structure in Cognitive Grammar. In Cognitive Grammar. Oxford: Oxford University Press.

ICLC-14 Book of Abstracts Theme sessions

EXPRESSIVE GGEMINATION IN THE

Nadežda Christopher SOAS [email protected]

Keywords: geminate, gemination, , double consonants, sound symbolism

The matter of geminates in the Russian language is a fairly straightforward one as far as books are concerned (Lopatin, 2009). Dvoinie soglasnie (double consonants) mainly occur due to affixation or compounding. However, a lot of these double consonants [CC] do not have the phonological extra length quality of a true geminate [C:]. In this paper I examine the previously undescribed Russian expressive geminates, and juxtapose them with the double consonants. I propose that the expressive geminates are the true geminates in Russian, arising from reduplication of phonological material. I also propose that further research can lead to the discovery of new linguistic universals, which expose the complex cognitive nature of many phonological phenomena cross- linguistically. Russian double (long) consonants are usually correspondent to two letters in , however, two consonants in writing do not always indicate a phonological double consonant, as the examples in (1) show. The Russian are followed by their transcription and meaning; stress is marked by the accent mark. (1) Россия – [rasíja] – ‘Russia’; кристалл – [kristál] – crystal; коррозия – [karózija] – ‘corrosion’). The examples in (2) contain words with double consonants that arose due to the morphological processes of affixation or compounding (the boundaries are indicated by a dash ‘-‘). (2) Без-закон-ный – [bezzakónnyj] – ‘unlawful’; глав-врач [glavvráč] – ‘head physician’. We can see here that the orthographic double consonants are reflected in the pronunciation of these words and are realised as [CC]. I argue that these are not the true Russian geminates as each of the two consonants in a cluster is pronounced somewhat separately, accentuating the orthographic double representation, whereas in the case of gemination, one consonantal sound is prolonged, and would be represented as [C:], as in the example in (3). (3) Original form Meaning Geminated form Meaning хороший ‘good’ [har:óshij] ‘very/extremely good!’; can also be used [haróshij] sarcastically to mean ‘not good at all’ or ‘silly’ (of a person)

As well as the change in the quality of the sound, we observe the change in meaning, which can be characterized as expressive colouring or intensification (verified with ten native speakers of Russian). The next step is to consider gemination as an instance of reduplication within the wider context of , which can be full or partial and predominantly has ‘an expressive connotation’ (Israeli, 1997:588). I propose that Russian expressive gemination is, in fact, an instance of full reduplication of a phonological segment (in our case, a consonant), which results in intensification of meaning. What is especially interesting about this observation and proposal is that in cases of expressive gemination, the meaning of the whole word is intensified as the result of reduplication of only one of its consonantal constituents, which leads to further questions regarding the underlying cognitive nature of consonants and their sound symbolic characteristics. Further investigation into the nature of the expressive geminates in the Russian and other languages can lead to discovering new cognitive phonological universals, which would not be restricted to a single realm of linguistic knowledge, but rather demonstrate their interconnectedness.

References

Israeli, A., 1997. Syntactic Reduplication in Russian: A Cooperative Principle Device in Dialogues. In: Journal of , 27, pp. 587-609 [online]. Available at: http://www.sciencedirect.com/science/article/pii/S0378216696000288 [Accessed 6 August 2016]. Lopatin, V.V. ed., 2009. Pravila Russkoi Orfografii i Punktuacii. Polnii Akademicheskii Spavochnik. Moscow: Eksmo. ICLC-14 Book of Abstracts Theme sessions

THEORY OF THE PHONEME IN THE RUSSIAN LINGUISTIC TRADITION

Natalia Kuznetsova Institute for Linguistic Studies, Russian Academy of Sciences [email protected]

Keywords: cognitive phonology, functionalism, theory of phoneme, sign systems, Russian linguistics

The of the phoneme was born inside the Russian-speaking linguistic tradition (Baudouin de Courtenay, Kruszewski, Scherba) and was directly related to the neurological science of the time. Baudouin de Courtenay defined the phoneme as a “homogenous, linguistically indivisible anthropophonic impression, emerging in the soul by the psychic merge of the impressions from pronunciations of the same sound” (1899: 355). Moreover, the word “psychic” appears to be used as synonymous to “cerebral” and “central nervous” in his writings (ibid.: 196, 354). His views were based on the contemporary revolutionary ideas by Sechenov that the central nervous system keeps the traces of the previous impressions. The more the impression is reiterated, the clearer the trace becomes and the longer it is kept by the nervous system (1866: 62). These ideas sound very modern and can be traced in contemporary works in the framework of functional (cognitive, usage-based) phonology (viz. Bybee 2001: 33, 52, Langacker 2008: 220). Later, the was divided into the so-called Scherba’s school (which followed these original anthropocentric ideas) and the more lingvocentric Moscow and Prague schools. Modern conceptions within the Scherba’s school will be discussed in the paper. They are based both on the functional theoretical ideas and the ongoing experimental research in , speech production and perception. For example, on the question of whether phonemes are basic operational mental categories (cf. Nathan 2006: 189), it is believed that “the codes of mostly supraphonemic levels are used” in speech perception, a phonemic string being a collateral product of speech recognition. However, the phonemic code remains available, otherwise both the language system and the perceptive mechanism would be deprived of openness (Zinder and Kasevič 1989: 36-37). In speech production and perception, both top-down and bottom-up analyses (from phonemes to morphological units and the reverse) are therefore used according to the needs. As for the nature of the phoneme, a coherent application of the phenomenological method, essential for cognitive linguistics (Langacker 2008: 31), brings Kuznetsova (2014) to the conclusion that phonemes are no less abstract symbolic units than other kinds of language signs. In the original semiotic conception by de Saussure the “signifiant”, an acoustic image of the word, was also an “entirely psychic” rather than physical phenomenon (1919: 29). For the phoneme, one can further distinguish the functional dimensions similar to those of other language signs: “semantics” (a structure of the mental acoustic image), “” (rules of distribution with other units of the same phonemic system under various conditions set by phonotactics, word prosody, morphology), and “pragmatics” (interfaces with any other entities outside the phonemic system in question, e.g. morphological units as wholes, graphic and orthographic correlates, phonemes of earlier diachronic stages or different regional variants of the same language, physiological restrictions on language production and perception). The whole set of criteria structured along these three dimensions allows to give an accurate description in fine functional details for any phoneme.

References

Baudouin de Courtenay [Boduen de Kurtene], Ivan A. 1963 (1899). Izbrannyje raboty po obš’emu jazykoznaniju, Volume. I. Мoscow: Izdatel’stvo AN SSSR. Bybee, Joan L. 2001. Phonology and language use. Cambridge: Cambridge University Press. Kuznetsova, Natalia V. 2014. Ob istorii, suš’nosti i izmerenijah fonemy. In Valentin F. Vydrin, Natalia V. Kuznetsova (eds.), Ot Bikina do Bambal’umy, iz var’ag v greki. Ekspedicionnyje et’udy v čest’ E.V. Perekhval’skoj. St.-Petersburg: Nestor-Istorija. 405-442. Langacker, Ronald W. 2008. Cognitive grammar: A basic introduction. Oxford: Oxford University Press. Nathan, Geoffrey. S. 2006. Is the phoneme usage-based? ― Some issues. In International Journal of English Studies 6 (2). 173–194. Saussure, Ferdinand de. 1997 (1919). Cours de linguistique générale. Paris: Payot & Rivages. Sečenov, Ivan M. 1866. Refleksy golovnogo mozga. St.-Petersburg: Tipografija imeni A. Golovačeva. Zinder, Lev R. and Vadim B. Kasevič. 1989. Fonema i eje mesto v sisteme jazyka i rečevoj dejatel’nosti. In Voprosy jazykoznanija 6, 29–38. ICLC-14 Book of Abstracts Theme sessions

PHONOLOGICAL AFFINITY AND SEMANTIC EXTENSIONS IN A NETWORK: GŌU WORD FAMILY IN CHINESE AS AN EXAMPLE

Chihkai Lin Soochow University, Taiwan [email protected]

Keywords: Chinese word family, network, phonological affinity, semantic extensions, metonymy,

This paper investigates gōu word family in Chinese from a network approach in cognitive linguistics. In Chinese, characters sharing similar components form a family. The word family must have a core, from which other words derive by adding other phonetic or semantic components. In a word family, the phonology shows related, but not necessarily identical, representations, and the meanings are also correlated via the extension of metonymy or metaphor. The phonological representations interact with the semantic extensions, creating a word family that connects the characters to form a network. While a word family forms a network, the issue concerning word families in Chinese is always given to Chinese paleography. The mechanism behind a word family in Chinese is seldom discussed in great detail in linguistics, let alone investigation from cognitive linguistics. In cognitive linguistics, phonology is seldom the mainstream, but still discussed in Bybee (2001), Mompean (2014), Nathan (2008, 2015), Taylor (2003), among others. This paper adopts a network approach (Bybee 2001, Taylor 2003, Uehara and Kumashiro 2007) dealing with connection between sounds and meanings. The network approach is comprised of three elements: schema, prototype and extension. Although the network approach in cognitive linguistics and word families in Chinese are similar in the mechanism of connecting sounds and meanings, there is no lengthy study focusing on how a word family in Chinese is interpreted by a network approach in cognitive linguistics. To investigate word families in Chinese from a network approach in cognitive phonology, this study concentrates on gōu word family in Chinese. This word family includes 50 Chinese characters, divided into two layers depending on whether the characters appear in the early dictionary, Shuōwén jiĕzì ‘explaining and analyzing words’ (SWJZ). The first layer includes 37 characters in SWJZ, and in the second layer there are 13 characters without any attestation in SWJZ. The gōu word family expresses a core meaning of newly-sprouted grass, and the phonological reconstruction of the core is *kug (Li 1972). In the first layer, the 37 characters show five phonological representations, *kug, *kjug, *gjug,*hug and *hjug, and the characters are divided into five semantic classes. Via metonymy, three classes are considered the semantic extensions of the core meaning: bending, tiny, and newly-born. Besides the three extensions, the other two classes are onomatopoeia and phonetic component, which are simply based on phonological affinity. In the second layer, there are four phonological representations for the 13 characters, which also show four meanings. Three of the four phonological representations follow those from the first layer, and one is an innovation. As for the meanings, only tiny, onomatopoeia and phonetic component from the first layer are retained in the second layer. A new meaning ‘fright’ based on metaphor is generated in the second layer by adding semantic components related to heart or eye. The network approach not only provides a detailed phonological affinity of the characters in gōu word family but also suggests a process of semantic extension in the word family: core meaning > metonymy > metaphor.

References

Bybee, J. 2001. Phonology and Language Use. Cambridge: Cambridge University Press. Mompean, J.A. 2014. Cognitive Linguistics and Phonology. In The Bloomsbury Companion to Cognitive Linguistics, J.R. Taylor & J. Littlemore (eds.), pp. 253-276. London: Bloomsbury Publishing. Nathan, G.S. 2008. Phonology: A Cognitive Grammar Introduction. Amsterdam: John Benjamins. Nathan, G. S. 2015. Phonology. In Handbook of Cognitive Linguistics, E. Dąbrowska & D. Divjak (eds.), pp. 253- 273. Berlin & New York: Mouton deGruyter. Taylor, J. R. 2003. Linguistics Categorization: Prototypes in Linguistic Theory. Oxford: Clarendon Press. Uehara, S. and F. Kumashiro. 2007. Onyin Keitai no mekanizumu: ninchi onyin keitairon no apuroochi [The Mechanism of Phonology and Morphology: An approach to cognitive phonology and morphology]. Tokyo: Kenkyusha. ICLC-14 Book of Abstracts Theme sessions

MOTIVATION IN PHONOLOGY: THE CASE OF /r/-SANDHI

Jose A. Mompean University of Murcia, Spain [email protected]

Keywords: motivation, phonology, /r/-sandhi, ‘linking’ /r/, ‘intrusive’ /r/

Motivation refers to the shaping of language structure by theory-independent, general cognitive processes as well as by bodily experience, phonetic processes, social/communicative interaction, cultural norms or even the internal ecology of the linguistic system itself. Motivation is a central topic in cognitive linguistics and it has been addressed in fields such as grammar and the lexicon (e.g. Panther & Radden, 2011; Radden & Panther, 2004). However, less attention has been paid to the motivation of phonological structure (but see Mompean, 2014 for an exception). The current paper addresses the topic of motivation in phonology within the framework of cognitive linguistics. Motivation is analysed with regard to the phenomenon of non-rhotic English /r/- sandhi, or the use of /r/ in-between vowels across to prevent hiatus. Following all previous accounts, two types of /r/-sandhi are discussed: ‘linking’ /r/ (e.g. more /mɔː/ but more and more

[mɔːɹ‿əm mɔː]) and ‘intrusive’ /r/ (e.g. saw [sɔː] but I saw it [aɪ sɔːɹ‿ɪt]). The paper discusses several aspects of the possible motivation of /r/-sandhi, including phonetic grounding (Pierrehumbert, 2000), usage-based aspects (Bybee, 2001), general cognitive processes such as categorization and schematization (Taylor, 2002) and lectal and cultural grounding (Kristiansen & Dirven, 2008). It assesses the importance of each of these sources of motivation in the shaping of /r/-sandhi with reference to previous studies (e.g. Hay & Sudbury, 2005; Mompean & Mompean-Guillamon, 2009; Pavlík, 2016; etc.) as well new data based on a longitudinal study of the speech of Queen Elizabeth II’s Christmas speeches. The findings of this study and others reveal that /r/-sandhi is shaped by different motivating factors including phonetic (e.g. prosodic structure), sociolinguistic (e.g. social status) and usage-based (e.g. collocation frequency) variables as well as idealized cognitive models of the relationship between spelling and pronunciation. Finally, the paper draws conclusions for the study of motivation in phonology such as the formalisation of motivating principles in phonological work or the need to provide multifactorial motivational explanations in phonology.

References

Bybee, Joan L. 2001. Phonology and Language Use. Cambridge: Cambridge University Press. Hay, Jennifer & Andrea Sudbury. 2005. How rhoticity became /r/-sandhi. Language 81(4). 799–823. Kristiansen, Gitte & René Dirven (eds). 2008. Cognitive . Language Variation, Cultural Models, Social Systems (Cognitive Linguistics Research [CLR] 39). Berlin/New York: Mouton de Gruyter. Mompean, Jose A. 2014. Cognitive linguistics and phonology. In John Taylor & Jeannette Littlemore (eds.), The Bloomsbury Companion to Cognitive Linguistics, 253–276. London: Bloomsbury Publishing. Mompean, Jose A. & Pilar Mompean-Guillamón. 2009. /r/-liaison in English: An empirical study. Cognitive Linguistics 20(4). 733–776. Panther, Klaus-Uwe & Günter Radden (eds.). 2011. Motivation in Grammar and the Lexicon. Amsterdam: John Benjamins. Pavlík, Radoslav. 2016. A usage-based account of /r/-liaison in Standard British English. Journal of 54. 109–122. Pierrehumbert, Janet B. 2000. The phonetic grounding of phonology. Les Cahiers de l'ICP, Bulletin de la Communication Parlée 5. 7–23. Radden, Gunter & Klaus-Uwe Panther (eds.). 2004. Studies in Linguistic Motivation (Cognitive Linguistics Research [CLR] 28). Berlin and New York: Mouton de Gruyter. Taylor, John R. 2002. Cognitive Grammar. Oxford Textbooks in Linguistics. Oxford: Oxford University Press

ICLC-14 Book of Abstracts Theme sessions

BLENDED FEET? NON-PROTOTYPICAL PHONOLOGICAL IN RUSSIAN

Tore Nesset UiT The Arctic University of Norway [email protected]

From the point of view of cognitive linguistics, rhythmic units such as prosodic feet are of particular importance, since they relate linguistic rhythm to rhythm in music and other domains, and thus underscore the relevance of a non-modular approach to language. On the basis of a detailed analysis of foot structure in Contemporary Standard Russian, in this paper I argue that (Taylor 2003) and Conceptual Integration (Fauconnier/Turner 2002) offer valuable contributions to prosodic phonology. It is claimed that in addition to prototypical feet (iambs and trochees), languages may posit a number of non-prototypical feet, which can be analyzed as blends based on the prototypical feet. Conventional wisdom has it that prosodic feet are binary (either iambs or trochees), and that a given language has one and only one type of foot. However, in recent years these assumptions have been challenged, and internally layered ternary feet that combine both iambic and trochaic elements have experienced a modest revival (Martínez-Paricio/Kager 2016). Does this mean that anything goes? Are there no restrictions on possible foot structures in language? I argue that Prototype Theory enables us to resolve the seemingly confusing picture. While languages normally posit a prototypical foot (iamb or trochee), the universal inventory of prosodic feet also includes non- prototypical phonological structures, such as internally layered ternary feet. Prototype Theory allows us to maintain the prominence of iambs and trochees in prosodic theory, but at the same time accommodates deviations from these prototypical structures. How are non-prototypical feet related to the prototypical structures? I propose that Conceptual Integration offers an insightful approach. In Contemporary Standard Russian, the stress system provides evidence for trochaic feet; insofar as stress arguably defaults to the penultimate syllable (Lavitskaya and Kabak 2014), the stress system can be analyzed in terms of a right-aligned syllabic trochee. However, facts about vowel reduction in Russian point toward iambic feet; in [məlʌdóvə] ‘young (gen.sg.masc.)’ the syllable immediately preceding the stressed syllable displays less reduction ([ʌ]) than other unstressed syllables that have schwa. In order to provide a unified analysis of stress and vowel reduction I propose an internally layered ternary foot ((σσ́)σ). The inner layer represents an iamb that facilitates a principled analysis of vowel reduction, while the outer layer is a trochaic structure that accommodates (default) stress placement. The internally layered ternary foot is analyzed as a blend, which combines information from an iambic and a trochaic input space. In addition, the blended foot involves emergent structure pertaining to the layered organization of the blended foot. While the application of blending to phonology is still in its infancy, my analysis suggests that phonolgoical concepts provide valuable insights about the way we think.

References

Fauconnier, G./M. Turner. (2002) The Way We Think. New York. Lavitskaya, Y./B. Kabak. (2014) Phonological default in the lexical stress system of Russian. Lingua 150: 363–385. Martínez-Paricio, V./R. Kager. (2015) The binary-to-ternary rhythmic continuum in stress typology.Phonology 32: 459–504 Taylor, J. (2003) Linguistic Categorization. Oxford. ICLC-14 Book of Abstracts Theme sessions

A COGNITIVE APPROACH TO PHONOLOGY: EVIDENCE FROM SIGNED AND SPOKEN LANGUAGES

Corrine Occhino Rochester Institute of Technology – Center on and Language [email protected]

Keywords: cognitive phonology; emergent structure; arbitrariness; embodiment; usage-based approaches

With few exceptions (Mompean, 2006; Nathan, 2008; Välimaa-Blum, 2005), cognitive linguistics has paid little attention to cognitive theories of phonology, due in part to the belief that phonological content consists of meaningless building blocks, different in kind from constructions in the grammar. However, if we accept that language structure is both non-compositional and emergent, there exists no a priori reason to consider phonological structure as categorically different. Phonological form emerges from the same semantically rich, socially and pragmatically contextualized usage-events as other linguistic form. Construing meaning broadly, including any and all which evoke associations, the existence and the influence of such phonological mappings have been observed across modalities. Recent research has shown, for example, that vowel quality can index affect (Eckert, 2009), featural properties of phonemes influence perceived speed and effectiveness of drugs (Abel & Glinert, 2008), and motivated phonological features facilitate sign recognition in children (Ormel, Knoors, Hermans, & Verhoeven, 2009). Current theories of phonology, in which phonemes are meaningless units, struggle to account for the presence of any meaningful phonological mappings, often marginalizing such data as anomalous. This paper introduces a theoretical framework built on usage-based and cognitive approaches (Bybee, 2001; Langacker, 2008) which accounts for the emergence of both motivated and arbitrary phonological structures. I suggest that arbitrariness is but one potential outcome in the emergence of phonological content. On one end of the spectrum, ‘meaningless forms’ arise only in cases when meaning associated with a given form becomes so attenuated, due to frequent usage, that a form is perceived as ‘arbitrary’, e.g. /p/ in English. I will show, however, that even ‘clear cut phonemes’ can exhibit vestiges of associations which can be resurrected under the right circumstances. On the other end of the spectrum, a second possible outcome arises when phonological content retains its meaningful associations, and these mappings are not lost in the schematization process. This maintenance of form- meaning mapping stems from both language internal analogization, and language external ‘grounding’. Such is the case with the phoneme ‘B-handshape’ (flat-hand) in ASL, which despite widespread usage across the lexic-on, still retains properties of flatness. Importantly, this framework helps to explain why phonological systems in signed and spoken languages, which on the surface seem to differ greatly in terms of motivated mappings, actually arise from the same domain-general cognitive processes, at work in both modalities. I describe how our construal of the world contributes to the maintenance of form- meaning associations, as our understanding of language is filtered through, and gained by way of embodied experience. Thus, embodiment of linguistic content and our ability to construe our interactions in the world, and with our bodies, directly influences our construal of linguistic form.

References

Abel, G. A., & Glinert, L. H. (2008). Chemotherapy as language: Sound symbolism in cancer medication names. Social Science & Medicine, 66(8), 1863–1869. Bybee, J. L. (2001). Phonology and language use. Cambridge; New York: Cambridge University Press. Eckert, P. (2009). Affect, Sound Symbolism, and Variation. University of Pennsylvania Working Papers in Linguistics, 15(2), 70–80. Langacker, R. W. (2008). Cognitive Grammar: A Basic Introduction. New York: Oxford University Press. Mompean Gonzalez, J. A. (2006). The Phoneme as a Basic-Level Category: Experimental Evidence from English. International Journal of English Studies, 6(2). Nathan, G. S. (2008). Phonology: a cognitive grammar introduction (Vol. 3). Philadelphia: John Benjamins. Ormel, E., Knoors, H., Hermans, D., & Verhoeven, L. (2009). The Role of Sign Phonology and Iconicity During Sign Processing: The Case of Deaf Children. Journal of Deaf Studies and Deaf , 14(4), 485–502. Välimaa-Blum, R. (2005). Cognitive phonology in : analytic tools for students of English. Berlin; New York: Mouton de Gruyter. ICLC-14 Book of Abstracts Theme sessions

PHONOLOGICAL FORMS AS PERCEPTUAL CATEGORIES: WHAT DO WE (NOT) KNOW?

Paul Olejarczuk, Vsevolod Kapatsinski University of Oregon [email protected], [email protected]

Keywords: category learning; sound categorization; distributional learning; novelty bias; phonology

While we know much about the internal structure of semantic categories (see Taylor, 1995, for a review), we know relatively little about phonological categories. Kapatsinski et al. (2016) examined learning of intonation contour categories and argued that these categories are fairly abstract, with little internal structure (cf. Nathan, 1986). Contrary to predictions of exemplar models, new exemplars were no more likely to be excluded from the category than familiar exemplars, as long as distance to the prototype was controlled. Contrary to predictions of prototype models, the prototype was no more acceptable than experienced exemplars or even novel exemplars, as long as they were just as far from the prototype as those experienced in training. These results were, however, consistent with window/boundary-based categorization (e.g. Keating, 1990). Olejarczuk & Kapatsinski (2016) exposed English speakers to novel categories of tones, defined by the magnitude of the pitch excursion comprising a LHL tone superimposed over /ka/. After training, participants were asked to judge the typicality of new and old exemplars situated along the pitch excursion continuum and to produce what they was the most typical /ka/. Both typicality and produced excursions were well predicted by the log exposure frequencies of training exemplars. The logarithmic transformation decreases the difference between the frequent and the rare while increasing the difference between the rare and the non-existent. Thus, raw frequency over-estimated the goodness of the most frequent stimuli and underestimated the goodness of those that were experienced but rare. Similarly, productions were affected by rare exemplars more than their raw frequency would predict. We show that, if production tracked typicality (which tracks log frequency), within-category frequency distributions would evolve towards the uniform distribution best fit by a window model. However, frequency distributions in production are not uniform, having well-defined modes (e.g. Koenig, 2001). Non-uniformity of production distributions must be explained by some pressure counteracting the salience of the novel in perception. Following Zipf (1949), we argue that this pressure comes from a rich-get-richer loop in production: the more one reaches for a production target, the easier it is to reach for it again. Unchecked, this pressure would reduce the distribution to a single spike in the middle. The frequency distributions we actually see are therefore a compromise between the perceptual pressure towards within-category uniformity and the production pressure towards singularity.

References

Kapatsinski, V., Olejarczuk, P., & Redford, M.A. (2016). Perceptual learning of intonation contour categories in adults and 9‐to 11‐year‐old children: Adults are more narrow‐minded. . DOI:10.1111/cogs.12345 Keating, P.A. (1990). The window model of coarticulation. In J. Kingston & M. E. Beckman (Eds.), Papers in Laboratory Phonology I, 451-470. Cambridge: Cambridge University Press. Koenig, L.L. (2001). Distributional characteristics of VOT in children's voiceless aspirated stops and interpretation of developmental trends. Journal of Speech, Language, and Hearing Research, 44(5), 1058-1068. Nathan, G.S. (1986). Phonemes as mental categories. Berkeley Linguistics Society, 12, 212-223. Olejarczuk, P., & Kapatsinski, V. (2016). Attention allocation in phonetic category learning. Proceedings of Cognitive Modeling in Linguistics, 14, 148-156. Taylor, J. R. (1995). Linguistic categorization. Oxford: Clarendon Press. Zipf, G. K. (1949). Human behavior and the principle of least effort. Oxford: Addison-Wesley.