Distinctive Features

Abstract

Distinctive feature theory is an effort to identify the phonetic dimensions that are important for lexical contrasts and phonological patterns in human languages. The set of features and its explanatory role have both expanded over the years, with features being used to define not only the contrasts but the groupings of sounds involved in rules and phonotactic restrictions, as well as the changes involved in rules. Distinctive features have been used to account for a wide range of phonological phenomena, and this chapter overviews the incremental steps by which the feature model has changed, along with some of the evidence for these steps. An important point is that many of the steps involve non‐obvious connections, something that is harder to see in hindsight. Recognizing that these steps are not obvious is important in order to see the insights that have been made in the history of distinctive feature theory, and to see that these claims are associated with differing degrees of evidence, despite often being assumed to be correct.

Opacity in Campidanian Sardinian metaphony

“…In the model proposed in this paper in which there is no one-to-one relationship between phonological representations and specific phonetic realizations, this kind of opaque chain shift can be accounted for. This assumption is in line with other recent studies acknowledging the existence of dissimilarities between phonology and phonetics; for a discussion on substance-freedom in phonology, that is, the fact that phonological computation is independent from phonetic interpretation, see Clements 2001;Blaho 2008;Odden 2006Odden , 2013Hale and Reiss 2008;Mielke 2008;Iosad 2012; 15 Given Richness of the Base, we should also consider the possibility of a high-mid input vowel specified as the derived high-mid vowel containing the features {front,high,low}. We included this potential input vowel; yet, the same set of four possible metaphonic patterns (as in 36) was derived.…”

Final devoicing and vowel lengthening in Friulian: A representational approach

“…This means that an explanation for coda-sonority facts could be sought in functional and/or historical factors rather than some universal property of synchronic phonological computation. This might be not very satisfying for those committed to explaining the extent of variation solely in terms of constraint reranking, but still remains a possibility (Kavitskaya, 2002;Blevins, 2005;Barnes, 2006;Mielke, 2007;Yu, 2007;Reiss, 2007). Second, issues related to the featural representations of sonority remain unresolved at this point.…”

“…Instead, features are emergent from generalizations based on actual data from a given language. Features are therefore assigned to segments on a language-specific basis, relying on overt evidence from that language rather than on a priori generalizations with respect to the behaviour of phonological primitives: for similar approaches and discussion, see Morén (2006Morén ( , 2007; Mielke (2007); Blaho (2008); Boersma and Hamann (2008);Boersma (2009); precursors to this line are found in structuralist thinking, e. g. Martinet (1955);Hjelmslev (1975). In practice, this means that phonetically similar segments in different languages are not guaranteed to have identical or even similar phonological representations, since the latter depend exclusively on phonological behaviour.…”

The Oxford Handbook of Language Evolution

“…This is demonstrated by Mielke (Mielke, 2008) after studying almost 600 of the world's languages, and previously argued by the eminent phonetician Ladefoged. MacNeilage points out that acoustic correlates of major speech features constitute continua, with no underlying discrete categories (as those working in speech recognition know to their cost).…”