PETER AUER a Note on Prosody in Natural Phonology
Total Page:16
File Type:pdf, Size:1020Kb
Sonderdrucke aus der Albert-Ludwigs-Universität Freiburg PETER AUER A note on prosody in natural phonology Originalbeitrag erschienen in: Julian Mendez Dosuma (Hrsg.): Naturalists at Krems: papers from the Workshop on Natural Phonology and Natural Morphology, (Krems, 1 - 7 July 1988). Salamanca: Ed. Univ. de Salamanca, 1990, S. 11-22 A Note on Prosody in Natural Phonology Peter Auer . Naturalness in Phonology: articulation/perception or cognition? The last 15 years have brought about dramatic changes in (Generative) Phonology; We have witnessed the times of segmental Generative Phonology coming to an end. Radically new conceptions of Phonology have developed. Instead of taking the input and output of Phonology as strings of linearly ordered segments, as it was the case in classical Generative Phonology, phonological units are now assumed to display an internal hierarchical organization; from year to year, the complexity of this internal organization seems to increase, and new research paradigms such as Autosegmental Phonology, Metrical Phonology, Prosodic Phonology, Dependency Phonology have developed and worked on phonological representations and models that have little to do with, say, SPE phonology. All of these new approaches make use of the theoretical insight that phonological data should be spread onto a number of levels or "tiers") This non-linear approach has made it possible to analyze suprasegmental phenomena, which by their very nature, spread beyond single phonological segments (such as tone, intonation, st0d, musical and pitch accent); in addition, phenomena hitherto considered to be in the realm of the segment turned out to be analyzable in a much more explanative way, when looked upon in the new paradigm - phenomena such as compensatory lengthening or assimilation. Indeed, the very nature of the phonological segment itself, which was long taken to consist of a bundle of features (a "matrix"), is today open to discussion and becomes gradually replaced by a hierarchical structure of features. 1 The idea is, of course, not as revolutionary as that if we take phonology as a whole; Firthian phonology was based on the same idea, and pre-structuralist phonologists of course always took into account prosodic features in explaining, for example, the development of the Germanic unaccented syllables. However, it was a revolutionary idea within generative phonology of the 'classical' kind. 12 Peter Auer The new trend in Generative Phonology has brought back a number of ideas and concepts from pre-generative, even pre-structuralist times, and it may not be exaggerating to see as one of its main assets, a new found tradition in phonology: Natural Phonology (NP), flourishing in the early seventies, has at least according to some (possibly) unbiased onlookers such as Bertinetto (1985: 581), played an important, though indirect role in bringing about the change from segmental to non- linear phonology, by insisting on the deficiencies of the SPE type of Generative Phonology and by bringing the syllable back into Phonology. It is all the more surprising that NP has profited very little from the progress that has been made since, and has had little to contribute to it. Instead, it is still restricted (with no exception. to my knowledge) to linear segmental analysis! Writers in the NP framework continue to cling to the "teleologies" of fortition and lenition as an overall and comprehensive explanation of phonological processes. Fortition (strengthening) and lenition (weakening), in turn, are understood as processes which enhance the perception or facilitate the articulation of particular segments (sounds). Both are intimately bound to the physiology of the speech and hearing organs. Other natural explanations for phonological phenomena than those that can be stated in terms of articulation (and. occasionally) perception, have played only a minor role. In fact, some authors seem to equate natural rules with processes that have their basis in the articulation system.' There is no theoretical reason why this should be so. Indeed, Stampe himself is much more dialectical on that point (although obviously arguing against the then up-to- date generative framework and not interested in prosody). He states his opinion somewhat paradoxically: processes are mental operations for him, not just peripheral; but these mental processes must have a "physical" phonetic basis; "It should not be supposed from this that processes are peripheral, physical events - merely the result of articulatory mistimings or of over- or under-shootings of articulatory targets. [...] Anticipatory substitutions, in particular, suggest that the substitutions occur in the central nervous system - i.e., that they are mental substitutions. The very suppressibility of processes argues for their mental nature. [...) But although processes are mental substitutions, they are substitutions which respond to physical phonetic difficulties." (Donegan & Stampe 1979: 136) Or, in the same line: "Although phonological substitution is a mental operation, it is clearly motivated by the physical character of speech - its neurophysiological, morphological, mechanical, temporal, and acoustic properties." (Stampe 1979: 6) 2 Cf. Vennemann (1986). 3 The few overview texts on NP do not mention prosody at all (Edwards & Shriberg 1983, Donegan & Stampe 1979, Dressler 1984). A passing remark can be found in Hurch (1988:18). 4 E.g., Hutcheson (1973: 1 f) equates natural processes with "ease of articulation" and Grammont's law "du moindre effort", Ralph (1981: 343) calls "natural rules" those that "are easily accounted for and formulated on the basis of substantive properties of the articulatory apparatus". and according to Wojcik (1981: 635) "the function of processes is to facilitate and/or clarify pronunciation", and "all processes have a teleology which stems precisely from this need to accomodate articulation or perception". The same equation between NP and articulatory phonetics can be found in some critical comments on NP; cf. Dressler's rejoinder (1984). 5 Note in passing that if the effects of phonological processes were restricted to peripheral neutral mechanisms it would not make sense to investigate aphasia as a testing ground for naturalness. A Note on Prosody in Natural Phonology 13 It is not easy to grasp what is meant by "physical" in these extracts, although the word is of key quality for their understanding. Stampe seems to allude to features of the articulatory organs ("mechanical"?), those of the acoustic signal ("acoustic", "morphological"?) and also to those of the (peripheral?) neural processing of language ("neurophysiological" seems to be opposed to "mental" here). Interestingly, he also alludes to the "temporal" character of language, but leaves it unclear if this is related to the acoustic, articulatory, or neural side of the "physics" of language. But this vagueness nonwithstanding, it can be followed with some justification that NP a la Stampe, although it has been almost exclusively concerned with articulatory explanation in praxi, is not theoretically restricted to this kind of explanation. Instead, it may well be the case that in addition to articulatory explanations of natural mental operations, there may be others. Here, I want to propose that some natural principles may have nothing to do at all with articulation or the physical properties of speech (whatever they may be); but are cognitive from the very start. This does not prevent the respective structures of a language obeying these principles from being "natural", in the sense of reflecting "the needs, capacities, and world of its users" (Donegan & Stampe 1979: 127). The basic tenet of all natural approaches to language, i.e. that it is necessary to distinguish between conventional, "symbolic" linguistic structures and processes on the one hand, and natural, language-independent ones on the other hand, does not lose its importance. But these "needs and capacities" are central neural processing strategies that apply to many, and also to linguistic, materials. This means that in order to catch up with what is going on in the rest of phonology, the very notion of naturalness must be conceived in a more comprehensive way in NP, which includes cognitive processes and structures. In addition to the well- known teleologies of articulatory weakening and perceptual strengthening, a process could then also be called natural if it conforms to general principles of cognitive processing. Note that this is the only way of conceptualizing naturalness that can also be applied to Natural Morphology and Natural Syntax. 6 Only if we see naturalness in a broader, cognitive sense, is it possible to develop a unified theory of naturalness comprising all levels of linguistic structure. Also note that phonology interacts in various ways with morphological and syntactic structure; and if we want to explain why many natural phonological processes have morphologically (or even syntactically) defined domains of application, we have to look at this interaction in terms of natural cognitive processing of meaningful units (morphemes, words, phrases). It is not difficult to find examples for the systematic blockage of otherwise natural phonological processes at morphological boundaries; one is the behaviour of geminates. It is well known that geminates, although being dominated by two timing positions (morae), usually behave like single segments, for they cannot be broken up by phonological processes, or by vowel epenthesis, even if