Other fronting movementsOne observation that links all these structures (including those with an initial
adverb) is that they are all pronounced with a pause after the initial phrase. This is
unlike the wh-element in spec CP, which has no pause after it. This might indicate, that
they are not in the same positions. Another observation that indicates a difference
between topicalisation and wh-movement is that there is never inversion with
topicalisation:
(102) a where will the robbers strike next
b *the bank, will the robbers strike next
Moreover, if the fronted topics occupied the same position as fronted wh-phrases,
then we would expect them to be in complementary distribution, which they are not:
(103) a on this train, where would you hide the money?
b this man, where have I seen before?
These data show us that the topic is not moved to the specifier of the CP, but to a
position to its left. The obvious suggestion is that the topic is adjoined to the CP. This
is supported by the fact that we can have multiple topics and adjunction is a recursive
structure:
(104) [CP yesterday, [CP on the train, [CP Biggs, [CP I saw]]]]
A complication is added by considering other examples. In embedded contexts, the
topic does not precede the CP, but follows both the specifier and the head
complementiser:
(105) a I asked where, in this town, we could hide
b I think that, in this town, there’s no hiding place
Thus it seems that there are two topic positions in the clause, one adjoined to the CP
and one adjoined to the IP. The choice of the two is not free however as it is only in
main clauses that the topic can adjoin to the CP and only in embedded clauses that the
topic can adjoin to IP:
(106) a where could [in this town], you hide?
b I asked [in this town] where you could hide?
Thus, the relevant structures for topicalisation in main and embedded clauses is as
follows:
(107) CP
PP CP
In this town where C'
C IP
can you hide