Although there is not extensive research on this issue, complementisers in Spanish follow the expected order and are found in early utterances either a little before or a little after wh-questions (see Barrena 2000).
The presence or absence of the complementiser in an embedded clause such as (I) often goes unnoticed and seems not to have any semantic or syntactic consequences.
The fact that the complementiser can be absent, as in (I)b, is not a general property of the language, as there are constructions where that cannot be absent:
Constructions (2) and (3) indicate that the complementiser is necessary and compulsory in certain contexts, as in subject relative clauses (2) and in finite subject clauses (3).
That is usually considered to be equivalent to the Spanish complementiser que, as it is the element that introduces finite embedded clauses after verbs like think, say or know, each taking a que in Spanish that cannot be omitted as illustrated in (4) below:
0] instead of [Spec,CP], thereby becoming derived complementisers, needs to be strongly supported, and it thus seems to me that the major differences between English restrictive and nonrestrictive relatives can be used as an argument in favour of such a claim.
In a parallel fashion to relative that, which has generally been considered a complementiser in transformational grammar (see the classical analysis in Bresnan ), restrictive who(m) and which, and likewise the null or empty operator, are argued in this approach to be eventually relative complementisers (see  above).
In section 3 above it was argued that the main feature distinguishing the present approach from other traditional frameworks is that restrictive who, whom, which, and also the null operator are taken as derived complementisers, since they are claimed to raise to the C head position.
The that-complementiser that introduces relative clauses differs in important ways from completive that, or from other complementisers in general.
The syntactic differences between retrictive and non-restrictive relative clauses that seem to be accounted for by treating the wh-operators in restrictive relatives as derived complementisers (i.