The ability to distinguish between individual objects and amorphous substances is a universal property of human cognition bound up with our perceptual apparatus, already present at very early stages of development. This universal cognitive ability finds linguistic expression in the nominal domain in the so-called “mass-count distinction”: nouns that denote substances (mass) have significantly different grammatical properties from those which denote objects (count). Chief amongst those properties is the distribution of plural morphology on mass and count nouns (apple vs. apples, mud vs. *muds).

However, this alignment between cognitive and linguistic properties appears to be imperfect. For example, recent research has shown that there are languages that appear to lack the mass-count distinction altogether. But the truly intriguing languages are those which do make the distinction but appear to buck the general trend concerning the ban on pluralisation on substance-denoting nouns. Greek is such a language. Tsoulas (2009) shows that although the language has a robust distinction between mass and count nouns, pluralising mass nouns is freely and systematically allowed: in Greek, waters may spill on the floor and electricities can be cut off.

As it turns out, Greek is not alone in allowing such patterns. A number of unrelated languages have been reported (albeit not extensively studied) to pattern in a similar way: Halkomelem Salish, Kuikuro, Innu-Aimun, Ojibwe, Blackfoot, Inuktitut, Telugu, Persian, and Turkish are particularly interesting cases. The phenomenon cannot therefore be attributed to idiosyncrasies of individual languages. Instead, it represents a point where languages may present structured variation that must be accounted for.

Taking advantage of the misalignment between cognition and perceptual experience on the one hand, and the grammar of natural languages on the other, we are presented with a unique opportunity to study linguistic variation in its purest form. However, and in contradistinction to more traditional areas of formal study of linguistic variation (such as word order, or the expression of subjects), this area has direct repercussions on foundational issues concerning the construction of meaning, semantic categorisation, and the relation of semantic theory to ontology over and above the study of formal linguistic properties.

The project is carried out at the University of York and at Ulster University and funded by the Leverhulme Trust.