In American Sign Language (ASL), pronominal functions are carried out by pointing signs: point to self for first-person, point to addressee or non-addressed persons or things as second- and third-person. This picture is complicated by the fact that pointing signs are also used in other ways: as one example, a signer can point to an abstract locus to refer to a referent associated with that locus. Analysis of these pointing signs has been controversial, starting with arguments that points are not really linguistic elements, just deictics (Johnston 2013). Even when considered as linguistic, there are numerous debates as to analysis. Meier (1990) argues that there is no 2nd/3rd grammatical distinction, but only a division between 1st and non-1st person; Koulidobrova & Lillo-Martin (2018) argue that some indexical points are demonstratives; Ahn (2020) argues that pointing to abstract locations serves as a modifier, and that co-speech pointing gestures can receive a similar analysis. In this presentation, I provide evidence from the acquisition of ASL and from co-speech pointing in young hearing children to contribute to the discussion of how pointing signs should be analyzed. The acquisition pattern shows first that the distribution of points to objects versus people is very different. Both signing and non-signing children produce points to objects from an early age (Lock et al. 1994). However, only signing children develop the use of points to self and other people before 3;00, highlighting the difference between gestural and linguistic pointing. Furthermore, the development of points to different persons supports the linguistic analysis, and provides potential support for the 2nd/3rd distinction. The results also point to an early development of demonstrative and locative signs, which can be differentiated from co-speech pointing. On the other hand, points to abstract loci are not used by either group of young children, suggesting that this is a different grammatical element.