Countability is a semantic feature related to the presence of a boundary (Jackendoff 1991); its opposite, uncountability, refers to the suppression of the reference to a boundary. Now, most (although possibly not all, as some of the papers in this volume argue) languages’ lexicon contain words which usually denote entities like bounded objects (chair), and others that denote entities whose boundaries are usually not pertinent–like substances (milk). 1 As compared to other semantic features, the importance of countability emerges from the fact that it involves the morpho-syntactic level. The close relation with the syntax can be observed by comparing how a same noun (cake), presented in a count syntactic context, entails a reference to a boundary and denotes an object (a cake), whereas in a mass syntactic context (some cake) the reference to a boundary is not encoded. At the morphological level, it is interesting to note that–in typologically different ways, but consistently across languages–countability constrains morphological Number, a nearly universally represented morphological feature (Corbett 2000; Dryer 2005). In fact, Number inflection is possible only for countable expressions, and uncountability allows only default morphological values to surface; for example, in most Indo-European languages, nouns used in mass references can only occur in the singular. The phenomenon of countability is pervasive through the structural levels of language, and its near universality uncovers its importance as a core grammar

1 In this regard, one could argue that there are pairs of words such as garlic vs. onions “where one is mass and the other is count and yet the items in the world that they describe seem to have no obvious difference that would account for this”(Pelletier 2012: 16). As clarified throughout the paper, we argue that boundaries and related concepts like shape are highly salient from a cognitive point of view (Prasada, Ferenz, and Haskell 2002) and likely to be encoded into language. Morpho-syntax, in particular, allows us to refer or not to the boundaries depending on the contextual relevance. Thus, since it is more probable that one cooks one onion, but just one garlic clove, it turns out that the word onion is mainly used in count contexts and so considered a count noun, whereas the word garlic is mainly used in mass contexts and so considered a mass noun.