This paper focuses on dynamic local cases, with special attention to spatial
descriptions which employ not only one but two or even several landmarks.
The study is based on a corpus of recorded Finnish dialect samples. The
data shows a strong goal bias, i.e., a tendency to favor expressions of goal
landmarks over sources, and especially so with descriptions with human
actors, as expected on the basis of previous studies. On the other hand, the
study does not directly support the idea that when the language has formal
means for differentiating between the expressions of source and goal, these
are aptly used to express more than one landmark per motion event. The data
is analyzed according to the verb type, the type of the spatial relationship
expressed by the landmark, and the construction type of the landmark
expression. It is shown that in Finnish, verbs of motion are more often than
other dynamic verbs accompanied by two or several landmark expressions,
which is an expected result. However, it is the construction type of the
landmark expression that gives the greatest motivation for the use of double
landmarks, which can be convincingly explained by the conceptualization
strategy of nesting.