The group of definable
automorphisms of a structure A is denoted by ℋ(A). The following theorem is used
to discover the group of definable automorphisms of various structures: If A has finite
type and A ≡ B then ℋ(A) ≡ℋ(B). It is also shown that every group may be
represented as the group of definable automorphisms of some structure.
Definable automorphisms are then investigated in infinitary languages. Finally
the notion of normal submodel is introduced in analogy to the notion of
normal subgroup with definable automorphisms playing the role of inner
automorphisms.
