Learning new words is a life-long skill that we maintain even through adulthood. A new word has two facets that differ both in their time course and their neural underpinnings. The lexical facet concerns the integration of the new word in the mental lexicon and its role in lexical dynamics, namely, competition during recognition. This facet emerges slowly, usually after a consolidation period involving sleep. The episodic facet relates to the temporal-spatial representation of the new word which becomes available immediately after encoding. The development of this facet usually benefits from the presence of semantic schemas, i.e., abstract mental structures to which new information can be anchored. However, the impact of semantic schemas on the development of the lexical facet is not as clear. Recent literature has provided mixed results, even suggesting the presence of schemas could be detrimental for lexicalization. The current project explores the role of semantic schemas on the development of the two facets of new spoken words (e.g., leoparna derived from the baseword leopardo and as such members of the same phonological cohort), by manipulating the visual information presented during encoding. Twenty four young adults heard the new words in a phoneme monitoring task on Day 1 and Day 2 (24h after the first session). Each new word was accompanied by a context of visual information presented in the center of the screen (within participants): in the orthographic context, its orthographic representation (e.g., leoparna); in the semantic context, a word (e.g., chávena; with no phonological or semantic association with the baseword) presented as a synonym for the new word. The two facets were evaluated in three sessions in order to study their time course: Day 1 (immediately after encoding); Day 2 (24h after encoding and before the second encoding phase); Day 8 (1 week after Day 1 and with no encoding period between Day 2 and 8). The lexical facet was indirectly studied through a semantic categorization task on the basewords (from which the new words were constructed), considering that if the new item (e.g., leoparna) had been lexicalized, participants would be slower to recognize the basewords (e.g., leopard), compared to control words (e.g., pérola, without any association with the new words but with the same psycholinguistic properties as the basewords). The lexical facet only emerged on Day 8 and it was only significant for words learned in the orthographic context: young adults were slower to categorize basewords associated to new words learned in the orthographic context compared to control words. To basewords related to new words learned in the semantic context no significant differences were observed in relation to control words, even after 1 week (although numerically there was an advantage for the categorization of control words on Day 8).These results suggest that the lexical facet is stronger for words learned in orthographic context, since they were integrated in the mental lexicon, influencing the recognition of words that were members of the same phonological cohort. The development of the episodic facet was immediate, in agreement with the previous literature: participants were able to recognize the new items and to recall them immediately after the first encoding phase, with this episodic facet becoming stronger throughout the sessions. However, unlike it was expected, the episodic facet of words learned in the orthographic context was stronger than the words learned in the semantic context. These results suggest that the presence of semantic schemas can interfere with the emergence of the lexical facet and, possibly, in the consolidation of the episodic facet of new words. Yet, if on one hand the interference observed for the emergence of the lexical facet can be explained by the complementary learning systems model McClelland et al., 1995), on the other hand, the interference observed on the consolidation of the episodic facet can be explained by a difficulty in the integration of the lexical information of the synonym-word and the new word. This possibility might suggest that semantic information access does not occur automatically but demands cognitive resources, at least when this information is presented in a written format. The current study converges with an arising body of literature that aims to fully comprehend learning new words as episodic and lexical representations.