The idea of a Multilingual Semantic Web is to provide access to knowledge available on the Semantic Web (SW) to speakers of different languages. In this chapter, we concentrate on a particular aspect of the Multilingual Semantic Web vision and discuss the challenge of multilingual verbalization for conceptual models. Natural verbalizations require n-ary fact types, whereas popular Description Logic (DL) dialects [such as those used by the Web Ontology Language (OWL)] only support binary fact types. We use a Fact-Based Modelling (FBM) approach because it supports n-ary verbalization. We also discuss the use of natural language taggers in the creation of these models, which preserves the natural form of those verbalization patterns. Patterns that are typical and representative are studied in English and Chinese. In order to publish such models to the SW, we use objectification (i.e. model reification) to transform n-ary fact types to a binary form.