An essential (or restrictive) relative clause is a referring expression that restricts the set of possible entities denoted by the head noun to a particular entity or set of entities that meet the criteria specified in this clause. A relative clause fulfills both of the following criteria:
- Taken as a whole, the clause functions as an attribute of the noun (N), answering the question What is the N like? or Which N?
- The structure contains a finite or nonfinite verb form distinct from the predicate of the main clause.[1]
Together, the relative clause and the nominal head it modifies constitute the relative construction.[2] Languages use a variety of relativizing strategies to mark relative clauses. Strategies worth consideration from a typological perspective are expressed morphosyntactically. The dominant[3] relativizing strategy of a language is the type employed by its subject-headed relative clauses.[4]
This parameter considers relative clauses in which the nominal head functions as the direct object of the relative clause.[5] Relative clauses of this type, known as object-headed relative clauses, may be expressed through a variety of morphological and syntactic strategies.
Types:
NoObjRel: The language does not have object-headed relative clauses.
ElseObjRel: Object-headed relative clauses are expressed only by means other than the standard relativizing strategy.[6]
RelPro: Object-headed relative clauses include both the nominal head of the relative clause in the main clause and an obligatory coreferential pronoun in the relative clause. The pronoun is marked as the object of the relative clause, while the nominal head is marked for the role it plays in the main clause.[7]
CorrelN: Object-headed relative clauses include both the nominal head in the relative clause and either the same noun or a more generic term in the main clause.[8]
CorrelPro: Object-headed relative clauses include both the nominal head in the relative clause, marked as the object, and a coreferential (personal or demonstrative) pronoun in the main clause, marked for the role it plays in the clause.[9]
RCO: Object-headed relative clauses include the nominal head in the relative clause, marked as the object; at the same time, the role of the ellipted nominal head in this clause is not marked. [10]
RCnonO: While the nominal head of a relative clause can be understood to be the object of the clause, this role is not marked grammatically; instead, the noun is only marked for its syntactic role in the main clause. [11]
When a language displays more than one type, multiple values can be listed. If one type is dominant, a slash (/) can separate the two values, with the dominant value appearing first; if neither is dominant, they are listed with an ampersand (&) separating the two. The use of parentheses indicates that the strategy is not obligatory.
[1] If the relative clause is removed from the sentence, the main clause should be able to stand alone as a grammatical clause.
[2] For example, in the construction the dog that barks, the nominal head is the dog, while the relative clause is that barks. In the participle-based example the barking dog, the nominal head is dog, while the relative clause is barking.
[3] The dominant strategy, for the purposes of this parameter, is that which occurs with greater frequency.
[4] Subject-headed relative clauses can be found in every language to date.
[5] The object of a relative clause is not necessarily the object of the main clause. Consider, for example, the following sentence: I spoke with the person whom you met yesterday. Although whom is the direct object of the relative clause, its nominal head, person, is the object of a preposition, not the verb, in the main clause. For a nonfinite relative clause, consider the sentence I spoke with the person mentioned by you yesterday.
[6] For example, a language might rely on a passive construction, in which the nominal head appears as the subject instead of the object, as in I spoke with the person who was mentioned by you. Alternatively, a language could use parataxis: the juxtaposition of two independent clauses without subordination or coordination, with neither clause containing any morphological features that refer to the other. If this value applies, the specific strategy should be detailed in the commentary.
[7] Languages of this type display the following structure: *The dog doesn’t bark that I see it. For a language to display this type, this structure must constitute a well-formed relative construction in that language.
[8] Languages of this type display the following structure: *Which dog I love, I am not afraid of the dog/animal. In this structure, the repetition of the nominal head is obligatory.
[9] Languages of this type display the following structure: * Which dog I see, it doesn’t bite or * Which dog I see, I wouldn’t give it a bone.
[10] This structure can be identified in sentences in which the nominal head is not the direct object of the main clause. Languages of this type display the following structure: *Whichever dogs I see I’m afraid. or *Whoever I’ve contacted I will gladly speak.
[11] This structure can be identified in sentences in which the nominal head is not the direct object of the main clause. Languages of this type display the following structure: *I love of the dog I am not afraid. In a nonfinite relative clause, it would appear as *The loved by me of the dog I am not afraid.