Information Bias Inside English Words*

Abstract

Previous works in cognitive science have reported that human cognition of words includes two preferences: a locational preference, by which word prefixes are remembered better than suffixes, and suffixes better than infixes; and a consonantal preference, by which consonants are remembered better than vowels. In this paper, the ambiguity with respect to prefix/infix/suffix and consonant/vowel is compared in terms of conditional entropy, by using large-scale data from English. The results show that consonants indeed have less ambiguity than vowels, and also, that the locational preference holds if word middles are considered as wholes.

How Linearity and Structural Complexity Interact and Affect the Recognition of Italian Derived Words

“…Prior morphological research has therefore generally involved treating these and other variables as continuous dimensions using multiple regression and multi-factorial analyses (Baayen et al 2007;Khorsi 2012; Kraska-Szlenk andŻygis 2012; Kuperman et al 2008Kuperman et al , 2009Kuperman et al , 2010Mirković et al 2011;Rueckl and Raveh 1999;Tanaka-Ishii 2012) or in categorical form using factorial designs (Assink et al 2000;Feldman and Larabee 2001;Kim et al 2015;Marslen-Wilson et al 1994;Meunier and Segui 2002;Taft and Ardasinski 2006). However, these perspectives are not mutually exclusive, since some researchers recognize the continuous or distributional nature of certain features, but still treat them categorically in their investigations (Burani and Thornton 2003;Burani et al 1997;Gonnerman et al 2007;Rueckl et al 1997;Vannest et al 2011); and still other investigators have used and reported both types of analyses as portions of the same overall study (Laudanna et al 1994;Marslen-Wilson et al 2008;Niswander-Klement and Pollatsek 2006).…”