Our prior research has found that infants have the ability to represent the numerosity of small collections of objects. In our speech segmentation project, we have incorporated this ability into a methodology that can be used to shed light on infants' processing of speech. A principal research question on the project concerns whether infants temporally segment speech into units, and if so, whether these units are phonemes or syllables. The findings are helping us resolve a theoretical dispute about early language development-whether infants initially use special-purpose speech processing mechanisms or whether they use general auditory mechanisms. This research project is also providing information about early numerical development by revealing part of the functional architecture of the infants' perceptual system, specifically a functional connection between the infants' ability to detect numerosity and a speech segmentation process.The project is funded by the National Institutes of Health. Prentice Starkey is the principal investigator.