This article may be too technical for most readers to understand. Please help improve it to make it understandable to non-experts, without removing the technical details. (September 2022) (Learn how and when to remove this template message)

This article includes a list of references, related reading, or external links, but its sources remain unclear because it lacks inline citations. Please help to improve this article by introducing more precise citations. (September 2022) (Learn how and when to remove this template message)

In linguistic semantics, an expression X is said to have **cumulative reference** if and only if the following holds: If X is true of both of *a* and *b*, then it is also true of the combination of *a* and *b*. Example: If two separate entities can be said to be "water", then combining them into one entity will yield more "water". If two separate entities can be said to be "a house", their combination cannot be said to be "a house". Hence, "water" has cumulative reference, while the expression "a house" does not. The plural form "houses", however, *does* have cumulative reference. If two (groups of) entities are both "houses", then their combination will still be "houses".

Cumulativity has proven relevant to the linguistic treatment of the mass/count distinction and for the characterization of grammatical telicity.

Formally, a cumulative predicate *CUM* can be defined as follows, where capital *X* is a variable over sets, *U* is the universe of discourse, *p* is a mereological part structure on *U*, and is the mereological sum operation.

In later work, Krifka has generalized the notion to *n*-ary predicates, based on the phenomenon of *cumulative quantification*. For example, the two following sentences appear to be equivalent:

- John ate an apple and Mary ate a pear.
- John and Mary ate an apple and a pear.

This shows that the relation "eat" is cumulative. In general, an *n*-ary predicate *R* is *cumulative* if and only if the following holds: