Traditionally, the verbal domain, as well as most parts of speech, has been divided into two broad categories, either functional or lexical. Functional verbals were verbs that were necessary as parts of speech to provide grammaticality to phrases (e.g. infinitive markers, auxiliaries, modals). Lexical verbs were verbals that provided purely semantic content to a phrase, meaning that if lacking, the content of the sentence would change (e.g. dance, sing, throw). However, there is previous evidence that within parts of speech, there are units that do not fall solely within either a functional or a lexical category. Littlefield (2006, 2009) proposed that the lexical and functional domains each have two dimensions, resulting in the following four-way division of parts of speech: [+lexical, -functional], [-lexical, -functional], [+lexical, +functional], [-lexical, +functional]. Using the CHILDES database (MacWhinney and Snow, 1985, 1990), Littlefield found evidence for each one of these semi-lexical categories in the prepositional domain by observing an observable difference in the order for which each division was acquired and mastered by children. Rubenstein and Martinek (2011), in a pilot study, found evidence for a similar four way categorical split through a similar order of acquisition for the verbal domain. I propose that using the CHILDES databases I can find further evidence for these four semi-lexical categories. I can furthermore find evidence for deeper, fine-grained categories within these domains, which would develop a spectrum of verbals ranking from least to most functional.