Markedness

Markedness is a necessary aspect of any theory of derivation. By definition, derivations have a starting point, and are necessarily ordered. To the greatest degree possible, this ordering should follow from conceptual necessity. The least marked elements in a structure are necessarily implemented last. In Nunes (2002) I show that there is evidence from a wide range of disorders, from the common case of ‘fronting’, saying key as TEA, and from a three and a half year old whose mother could barely understand a word she said, who had an extremely complex way of saying ‘simple’ one syllable words, with a most uncommon three way pattern of assimilation between the critical features differentiating pea, key, and tea, the lip action in P, the back of the tongue action in K, and the tip of the tongue action in T, that these features are implemented in that order. In the case of fronting, the middle step in the derivation is missed out. Across this wide range of disorders, the role of markedness is respected. But other aspects of the derivational sequence can be misconfigured.

There are degrees of markedness. Take the sound structure of a word like strange, with a cluster of  three phonemes in the onset and a cluster with N and the affricate represented by GE. Such a word represents a high degree of markedness in its sound structure. Conversely, the CV structures of Mummy and Daddy or Mama and Dada and so on, are minimally marked.

Similar or identical consonants occur in the overwhelming majority of languages.

There are many different conceptions of markedness, originating from Nikolai Trubetskoy in the 1930s, though he didn’t actually use the term, then developed by Roman Jakobson, Morris Halle, Noam Chomsky in the 1950s and 60s, and then many others. But the esseence of markedness is a measure of how much information has to be added to a structure in the course of its derivation. Unmarked structure represents what can be implemented by default.