Mixed emotions remain a fascinating, yet still understudied phenomenon. All of the previous research has focused solely on ambivalence, studying only the mix of positivity and negativity in emotions (the dimensions of valence). We sum up the already existing knowledge about the dimensional approach to ambivalence and its consequences. Based directly on this knowledge, we introduce a new theoretical model describing ambiguity in four additional dimensions (apart from valence), grouped into two bivariate spaces: origin (dimensions of automaticity and reflectiveness) and activation (arousal and subjective significance). Both of these spaces have never been studied before in the context of ambiguity and mixed feelings. Future implications of the new model are discussed, including any potential impact on the methodology of research and the possible advantages in understanding and describing emotional experiences.