The ability to understand simple vocabulary, conveyed by haptics, has demonstrated encouraging results, even with limited training. However, the level of achievable accuracy on communication indicative of real-world conversations, typically characterized by long phrases formed by complex vocabulary, remains unknown. This work presents an in-depth case-study analysis of the learning path of one experienced participant over a two-month period of daily practice with our vibrotactile communication apparatus. By the end of the two months, this participant was able to correctly identify 97% of the words in phrases consisting of 6-8 words, delivered as a sequence of vibration patterns. To the best of our knowledge, this work demonstrates the most advanced haptic rendering of language through a wearable device to date.