The December 2010 issue of 'Developmental Review' features a nice meta-analysis of of studies on pointing and language development by Cristina Colonnesia, Geert Jan J.M. Stamsa, Irene Kostera, and Marc J. Noomb. Here's their abstract and their 'research highlights'
The use of the pointing gesture is one of the first ways to communicate with the world. This gesture emerges before the second year of life and it is assumed to be the first form of intentional communication. This meta-analysis examined the concurrent and longitudinal relation between pointing and the emergence of language. Twenty-five studies were included into the meta-analysis, including 734 children. The role of several moderators was examined: pointing modality, pointing motive, age at which the pointing was measured, the assessment method of the pointing gesture and language development, the modality of language, SES, and country. The results showed both a concurrent (r = .52) and a longitudinal (r = .35) relation between pointing and language development.
The relation between pointing and language development became stronger with age, and was found for pointing with a declarative and general motive, but not for pointing with an imperative motive. It is concluded that the pointing gesture is a key joint-attention behavior involved in the acquisition of language.
► Pointing gesture is concurrently (r = .52) and longitudinally (r = .35) related to language development. ► The relation between pointing gesture and language becomes stronger with age, in particular during the end of the second year of life. ► Both the production (r = .33) and the comprehension (r = .38) of pointing gesture are related to language development. ► Pointing with a declarative (r = .39) and with a general motive (r = .39), rather than with an imperative motive (r = .04), are related to language.