ISA meets Lara: An incremental word space model for cognitively plausible simulations of semantic learning