In the past decades, neuroscientists and clinicians have collected a considerable amount of
data and drastically increased our knowledge about the mapping of language in the brain. The emerging
picture from the accumulated knowledge is that there are complex and combinatorial relationships between
language functions and anatomical brain regions. Understanding the underlying principles of this
complex mapping is of paramount importance for the identification of the brain signature of language
and Neuro-Clinical signatures that explain language impairments and predict language recovery after
stroke. We review recent attempts to addresses this question of language-brain mapping. We introduce
the different concepts of mapping (from diffeomorphic one-to-one mapping to many-to-many mapping).
We build those different forms of mapping to derive a theoretical framework where the current
principles of brain architectures including redundancy, degeneracy, pluri-potentiality and bow-tie network
are described.