Relative Clauses in Syntax
Summary and Keywords
A relative clause is a clausal modifier that relates to a constituent of the sentence, typically a noun phrase. This is the antecedent or “head” of the relative construction. What makes the configuration special is that the subordinate clause contains a variable that is bound by the head. For instance, in the English sentence Peter recited a poem that Anne liked, the object of the embedded verb liked is relativized. In this example, the relative clause is a restrictive property, and the possible reference of a poem is narrowed to poems that Anne likes. However, it is also possible to construct a relative clause non-restrictively. If the example is changed to Peter recited this poem by Keats, which Anne likes, the relative clause provides additional information about the antecedent, and the internal variable, here spelled out by the relative pronoun which, is necessarily coreferential with the antecedent.
Almost all languages make use of (restrictive) relative constructions in one way or another. Various strategies of building relative clauses have been distinguished, which correlate at least partially with particular properties of languages, including word order patterns and the availability of certain pronouns. Relative clauses can follow or precede the head, or even include the head. Some languages make use of relative pronouns, while others use resumptive pronouns, or simply leave the relativized argument unpronounced in the subordinate clause. Furthermore, there is cross-linguistic variation in the range of syntactic functions that can be relativized. Notably, more than one type of relative clause can be present in one language. Special types of relative constructions include free relatives (with an implied pronominal antecedent), cleft constructions, and correlatives.
There is an extensive literature on the structural analysis of relative constructions. Questions that are debated include: How can different subtypes be distinguished? How does the internal variable relate to the antecedent? How can reconstruction and anti-reconstruction effects be explained? At what structural level is the relative clause attached to the antecedent or the matrix clause?
Keywords: appositive relative clause, correlative, cleft, free relative, internally headed relative clause, modification, restrictive relative clause, operator construction, relative pronoun, resumptive pronoun
1. Main Characteristics of Relative Clauses
Relative constructions enable speakers to use a complex expression within another, larger expression. In this way, relativization embodies the recursive power of language par excellence.
Traditionally, a relative clause is described as a clausal modifier of a noun phrase, the antecedent. For a deeper understanding, a more precise characterization is required. In (1), the complex noun phrase between brackets is the internal argument of the predicate read and functions as the object of the main clause. It consists of an outer determiner (underlined), a relativized NP (in bold-face), and a relative clause (italicized). The meaning of the sentence is roughly specified in (1b).
The relativized constituent, although technically a phrase (potentially), is often referred to as the relative head or the head noun of the relative construction. Crucially, it does not only combine as a property with the outer determiner to form a referential expression, it also binds a variable within the embedded clause (here, in the internal argument position of published). This is a defining property of the relative construction. Notably, in a restrictive relative construction, the outer determiner takes as a restrictive term the combination of both the head noun and the relative clause, implying that it is hierarchically higher.
Relative clauses must be distinguished from various other types of embedded clauses. In (2a), a relative clause (RC) modifies the object of the main clause; in this case it specifies which woman it is that Peter admired. The RC itself contains a subject variable. In (2b) and (2c), there is no relativization: in the former, the italicized embedded clause itself is the object of the main clause; in the latter, the embedded clause functions as an adverbial. Both clauses are internally complete.
Even within complex noun phrases there is a difference between RCs and complement clauses, as is illustrated by the minimal pair in (3). Again, the RC in (3a) contains a variable that is bound by the head (Anne told x, and x is a rumor), whereas the complement clause in (3b) is internally complete. In effect, the RC serves to identify the referent of the complex NP (which rumor, as opposed to potential other ones?) by attributing another property to it. By contrast, the complement clause specifies what NP stands for (here, it reveals the content of the rumor).
There are also cases where the head is tacit. The so-called free or headless relative construction in (4a) can be understood as the thing that Anne had bought. The pronoun what is said to contain an implicit antecedent. Given rather common syncretisms in the pronominal domain, this can lead to confusion with embedded questions as in (4b).
It depends on the lexical semantics of the matrix predicate which interpretations—with associated syntactic configurations—are possible.
In principle, the syntactic function and theta role of the noun phrase containing a RC and those of the variable inside the RC are independent. In (5) all four permutations of subject and object are illustrated. On a par with commonly used terminology, “subject/object relative” here means a RC of which the subject/object has been relativized.
Every noun phrase, whatever its function in the matrix, can contain a relative clause. However, the internal role of the variable is not equally free. There appears to be an accessibility hierarchy,1 which roughly translates to the following hierarchy of syntactic functions: SU > DO > IO > Prepositional > Possessive > Adverbial > Comparative > Conjunct. If a language allows for, say, possessive relatives, all higher functions can also be relativized, but not necessarily the lower ones. The examples in (6) show that most functions are accessible in English, but not all. The examples in (7) show that the syntactic function of complex noun phrases in the matrix is unrestricted.
Psycholinguistic measurements have also shown that subject relatives are usually easier to process than RCs with lower functions, though this may not be universally the case.2
2. Cross-Linguistic Typology and Parametric Variation: A Concise Overview
Relativization is a universally available process, but it comes in various shapes. This section addresses some relevant aspects of the attested cross-linguistic variation, taking a parametric perspective. Important subtypes of relative constructions are highlighted further in section 3.
One striking property in which relative constructions differ is the position of the RC with respect to the head: it can be postnominal (as in English) or prenominal. This is illustrated with different languages in (8), where the heads are boldfaced and the RCs are put between brackets for expository purposes. In (8a), RP indicates a relative pronoun. In (8b), cited from Lehmann (1984, p. 64),3 the case particle bă is prenominal (i.e., it relates to the book); the particle de is a nominalizer with an attributive function.
There are no strong correlations between basic word order and the type of relative clause a language has. Although prenominal relative clauses tend to occur in OV languages, there are exceptions, and the reverse is not true: postnominal relatives are attested in both VO and OV languages.4
Interestingly, a relative head NP can also surface inside a RC in some languages. This is the head-internal or circumnominal pattern; see (9), from Lehmann (1984, p. 117). Here, nə is a subordinating particle.
It appears that this is hardly ever the only available relative strategy in a language. Dagbani, for example, also has postnominal relatives.
Usually, a relative construction (the head plus the RC) functions as a noun phrase in its context. Thus, the RC is structurally embedded inside an NP/DP. A systematic exception to this is the correlative construction illustrated in (10), from Lehmann (1984, p. 135). It involves a preposed clause that functions as an RC with an internal (normally fronted) head that is anaphorically resumed in the matrix. In this example, the resumptive element is a demonstrative.
As is the case for circumnominal relatives, the correlative construction can be the principal relativization strategy in a particular language, but it usually coexists with one or more other strategies.
A variant of the regular headed relative construction is the free or headless relative, in which the head noun is completely suppressed. This gives rise to a default definite or universal interpretation:
Here, the relative pronoun can be interpreted as the person who or every thing that, for instance.
Many Indo-European languages make use of relative pronouns. These serve both to introduce a relative clause as an attribute and to represent the RC-internal function of the head. Since a relative clause is an embedded clause, it can alternatively contain a clause-initial or final complementizer. In some cases, neither is found. All three possibilities can be illustrated in English:
Examples like (12c), without any linker, are sometimes called reduced relative clauses.6
In the absence of a relative pronoun, the internal function of the head can be expressed by a resumptive pronoun, as in (13), from Givón (1984, p. 655). Here, the relative element she is a complementizer; the relevant in situ pronoun is ota.
Generally, relative elements can express various functions: (i) subordination, (ii) an attributive link with the head NP by means of φ-feature agreement and a clause-peripheral position, (iii) a link with the RC-internal function via the expression of case and/or an in situ position.7
Relative pronouns occur in postnominal and correlative RCs only. They combine functions (ii) and (iii), often in an underspecified way (which makes classification difficult). Many involve a w(h)/q-morpheme like interrogative pronouns, which is consistent with preposing. Examples are English who and Serbo-Croatian koje. Some have a demonstrative core, like Dutch die or Danish den. Some consist of or contain a morpheme specialized to relativization, like Hindi jo or Slovenian kdòr. A final case involves the use of a classifier as a relative marker; this is illustrated in (14), from Lehmann (1984, p. 102):
If there is no sign of D-related features (a paradigm of φ-features, case, wh, or dem), a peripheral relative element is likely a Complementizer. Many of these simply equal a regular subordinating C like Norwegian som or Farsi ke. Some, however, are specialized for relative contexts, such as Czech co. Nominalizing particles, like Chinese de, are also used in relative constructions; these are normally clause-final.
A final class of relative elements consists of relative affixes on the finite verb inside the RC. These, too, come in various kinds. Either they are additional morphemes of an attributive, subordinative, nominalizing, or specialized relative sort or they replace a tense or agreement morpheme with a more specialized form. An example is (15), from Lehmann (1984, p. 78):
Here, aʔu is an additional relativizing morpheme.
Despite all these possibilities, some languages do not mark RCs in any way. An example is (16), from Lehmann (1984, p. 83), where both a subject interpretation and an object interpretation are possible.
Note that this is different from the situation in English, where zero relatives as in (12c) are optional, and used only in certain configurations.
The finiteness of the RC is also subject to cross-linguistic variation. An example of a non-finite RC is (17), from Lehmann (1984, p. 54), which is prenominal.
Participial relatives tend to be prenominal, but this is not universally the case. Also, the English construction a man to fear can be considered a postnominal non-finite relative.
Finally, consider definiteness and case. If an outer determiner is used, it may surface in various positions with respect to the head NP and the RC. A minimal pair is in (18):
A definiteness marker can also follow the RC, as in (19), from Lehmann (1984, p. 95). Compare also the RC-final plural marking in (15) above.
Similarly, case marking of the relative construction is variable, depending on the particular language. Two examples are in (20); (20b) is from Lehmann (1984, p. 79):
It is quite common to mark case on the head NP (and/or the external determiner), as in (20a), but (20b) shows that the RC as a whole can also receive a case morpheme. The case corresponding to the internal role of the head can be expressed on a relative or resumptive pronoun, if available, for example der in (20a). Importantly, the external and internal syntactic role of the head are—in principle—independent (recall the examples in section 1). Somewhat confusingly, there are attested instances of unexpected case matching between the head NP and a relative pronoun, for example in ancient Greek and Latin. Relative case attraction (attractio relativi) is the situation where the relative pronoun takes over the case assigned to the antecedent in the matrix; inverse case attraction (attractio inversa) is the reverse process where the antecedent takes over the case assigned to the pronoun RC-internally.8
All in all, there is a great deal of cross-linguistic and intra-linguistic variation in the morphosyntactic appearance of relative constructions, as might be expected.
3. Subtypes of Relative Constructions, and Issues Concerning Their Analysis
What follows is a catalogue of different classes of RCs and closely related constructions, each with a brief description of important properties and related theoretical discussions. Section 4 continues with analytical considerations.
3.1 Postnominal Headed Relative Clauses
Headed relatives of the postnominal variant are cross-linguistically widespread and have been subject to intensive syntactic research, especially since the 1960s.
Traditionally, RCs are considered as adjuncts to the antecedent. However, this is not exactly true for restrictive RCs, because they are clearly within the scope of determiners and quantifiers related to the head. For instance, (21) does not entail that “Peter loved all the women” or that “all the women left”:
This implies that an outer determiner does not form a constituent with the head noun: *[[ D N] RC]. The hierarchical structure must therefore be [ D [ N RC]], very roughly, and the term “relative head (NP)” is to be understood as excluding outer determiners/quantifiers. In a more elaborate structure, the RC can still be analyzed as an adjunct to NP or N’, or as a complement of N, but within a DP shell: [DP D [NP N(P) RC]].9 A more intricate alternative is the idea that the relative clause is in fact the complement of D, and somehow includes the preposed head: [DP D [CP NP . . .]]; see section 4.1 for more discussion.
Relative clauses containing a relative pronoun are Aʹ-movement constructions.10 The examples in (22) show that movement is unbounded (potentially successive cyclic) but sensitive to strong islands: who in (22c), for instance, crosses a complex NP boundary, which is ungrammatical.11
Since the same constraints apply in the absence of a visible relative pronoun, it is assumed that that- and zero relatives involve the exact same configuration, but abstractly so. In (23), OP represents an empty operator or an elided copy of the head.
A relative pronoun/operator instantiates the relative gap, the internal function related to the head NP. As such, it captures the internal theta role and case, as resumptive pronouns do. Furthermore, it is obligatorily linked to the visible external head NP. Fronting is one step in this process, but that in itself may be insufficient. Some sort of referential dependency or co-indexing under c‑command is a theoretical possibility:
Notice, however, that the head, contrary to the RP/OP, is not a full DP, and is in fact not referential: semantically, it expresses a property that is intersected with the RC that contains the variable. There is no generally accepted solution to this puzzle (but see section 4.1).
The example in (25) illustrates again that a relative head has in fact a phrasal status (deliberate attack on a city). Moreover, it shows that there can be pied piping of a larger phrase containing a relative pronoun, here from whose consequences.12
In many languages, a relative clause can be extraposed to the right, like certain other modifiers and major constituents, mainly PPs and CPs. This is illustrated in (26). Here, the RC is extraposed across the participle, which marks the right edge of the clause in Dutch. Resultingly, it is separated from the head. Note that extraposed restrictive material is included in the intonational contour of its matrix.13
Extraposition is clause-bound,14 so an RC extraposed from within a subject clause or any other non-peripheral embedded clause ends up in a clause-final but sentence-medial position:
An intriguing phenomenon that seems to build on the possibility of coordination and extraposition are split-antecedent RCs, as in (29):15
Such cases require an overt relative pronoun in English. See also section 3.6 on correlatives.
3.2 Non-Restrictive/Appositive Versus Restrictive Relative Clauses
Non-restrictive or appositive RCs (ARCs) add information about the antecedent but do not restrict its reference. The antecedent can be unique, definite, generic, or specific indefinite, but normally not non-specific:
Semantically, the relative pronoun is constructed as coreferential with the entire antecedent; hence the antecedent cannot be non-referential. This contrasts with the situation for RRCs: a noun phrase containing a RRC can be non-specific/non-referential in its context; compare (31a) to (31b), where negation induces non-referentiality of the direct object:
However, in certain contexts involving modal subordination (where a piece of discourse is interpreted as conditional on a scenario described in a previous one), the antecedent of an ARC can be considered referential within a possible world, and the relative pronoun behaves as an E‑type pronoun:16
In appositive constructions, it is clear that outer determiners/quantifiers do not outscope an ARC. For this reason, the syntactic configuration must be different from the one in restrictive relative constructions. Specifically, ARCs have a different site of attachment than RRCs. Roughly, the difference appears to be this: [ D [ N RRC]] versus [[ D N] ARC]. See section 4.2 for more discussion about the syntactic position of ARCs, and their resemblance to parentheticals.
Internally, ARCs are structured similarly to RRCs. The V-final word order in Dutch clearly shows that both are subordinate clauses. Also, the same relative pronouns can be used. The example in (33) is therefore ambiguous, and the intended reading must be provided by the context and the intonational pattern. As a rule, ARCs receive a comma intonation or, more precisely, a phonological Intonational Phrase (ι), unlike RRCs.17
In English, an ARC is normally introduced by an overt relative pronoun (although some speakers accept that-ARCs), but other languages (e.g., French, Danish) do allow for ARCs constructed with just a complementizer. A zero strategy for postnominal ARCs has not been attested (but see section 3.4).
For evident reasons, a restrictive RC cannot be attached to a proper name, unless this name can be interpreted as non-unique, hence restrictable, in the relevant context:
In this respect, RCs are no different from PP-modifiers: cf. the London of the past, etc.
Like restrictive RCs, appositive RCs can be extraposed, as in (36a), pending certain proximity constraints. Especially ARCs that receive a continuative or resultative reading are likely to be so (36b).18
If the intonational separation between the main clause and an extraposed ARC appears to be a full sentence division, the construction is called a relative junction (Fr. relatif de liaison). In such cases, the relative pronoun is usually (but not necessarily) a wh-pronoun that selects an event, an entire proposition or even a larger piece of discourse as its antecedent:
More generally, ARCs can take non-nominal antecedents, provided an appropriate context; see (38).19 In certain cases, a nominal retention of the antecedent is acceptable, for example fact in (38b).
None of this appears to be possible in restrictive relative constructions (but see section 3.7 concerning head retention).
3.3 Amount Relatives and the Like
Certain relative constructions receive a special interpretation in that they involve quantification over a variable that is not an individual variable but a variable over degrees, amounts, kinds, or events.20 Examples are provided in (39):
All of these involve a semantic maximalization operation, they cannot be combined with a (weak) indefinite external determiner, and stacking is impossible. For example, (39a) can be paraphrased as “There were exactly four birds in the cage, and we observed all of them.” The four birds cannot be replaced by an indefinite expression like a bird or some birds. Combination with an additional restrictive RC is possible, but stacking with another degree relative is impossible: the four birds that there were in the cage that (*there) had been flying freely before.
3.4 Prenominal Relative Clauses
A Basque example of a prenominal relative clause is given in (40), from Lehmann (1984, p. 59); see also (17) above, from Turkish:
Relative pronouns are not attested in prenominal constructions. Relative complementizers or particles are normally clause-final and do not equal the regular embedding complementizer.21
The status of potential appositive prenominal relatives is debated. To the extent that prenominal RCs can be interpreted non-restrictively, they seem to behave like decorative or non-intersective adjectives.22
3.5 Internally Headed Relative Clauses
Internally headed or circumnominal RCs are nominalized clauses containing a noun phrase that is interpreted as the head of a relative construction. The IHRC as a whole surfaces in a noun phrase position within the matrix. Internal relative pronouns (let alone resumptives) are excluded. An illustration from Quechua is given in (41), from Cole (1987, p. 277); see also (9) above, from Dagbani:
The (ambiguous) example in (42), from Lehmann (1984, p. 111), shows that there can be external determiners and case markers, which, if present, are always right-peripheral.
If the head NP happens to surface RC-peripherally, it requires further investigation to determine if the RC construction is actually internally headed, or in fact pre- or postnominal instead.23
There appear to be two different main types of IHRCs, along the lines of a semantic split: IHRCs are restrictive in some languages (e.g., Lakhota, Mojave) but obligatorily maximalizing in others (e.g., Quechua, Japanese, Korean).24 Restrictive IHRCs allow for an indefinite reading (i.e., existential quantification); maximalizing ones do not, and resist stacking (cf. section 3.3). In maximalizing IHRCs, the internal head position may exhibit strong determiners (roughly, definites and universal quantifiers), contrary to the situation in restrictive ones, where only an external determiner can be strong.
The syntactic and semantic derivation of IHRCs is debated. A major question is how the internal noun phrase can be interpreted as the relative head, and hence be related to the nominalized clause as a whole (and/or a potential external D‑position)—the linking problem. Covert Aʹ-movement of an operator or even the entire internal head is an imaginable solution, but the reported lack of island effects in some languages suggest a more conservative approach in terms of scope marking or an E-type dependency. It seems likely, however, that the two different main types of IHRCs require a somewhat different analysis.
Correlatives are also internally headed. They are in a left-peripheral position with respect to the matrix, which itself contains a resumptive element (also called the correlate): a personal or demonstrative pronoun, possibly even a retention of the relative head. An illustration from Hindi is given in (43), from Grosu and Landman (1998, p. 164); see also (10) above, from Bambara.
Unlike IHRCs, correlative clauses are not nominalized: external determiners, case markers, and adpositions are excluded. On the other hand, an internal relative pronoun seems obligatory.
Correlatives have been shown to be maximalizing, hence the resumptive is necessarily definite or universal (i.e., weak determiners cannot be used); moreover, stacking of correlatives is impossible.25
There are attested locality constraints on the hierarchical distance between correlative clauses and the resumptive in the matrix, as well as reconstruction effects (such that the head noun behaves as if it occupies the position of the correlate in terms of scope; see also section 4.1). This suggests that correlatives are moved from a base-position next to the correlate to their left-peripheral position within the matrix.26 If so, instances of multiple relativization like (44), slightly adapted from Bhatt (2003, p. 492), which contain more than one different head noun, must be treated as exceptional constructions.
Such possibilities are cross-linguistically very restricted. A somewhat related problem seems to be the (marked) option of a split antecedent, which can be found in English:27
Here, the extraposed relative clause cannot easily be reconstructed as a postnominal RC belonging to one or each of the two antecedents separately.
3.7 Double-Headed Relative Clauses
It has been reported for various languages that children acquiring RCs may repeat an external head inside a RRC, as in (46), from Pérez-Leroux (1995, p. 114). In effect, this results in a blend between externally and internally headed RCs:
Notably, a full resumptive DP inside a RRC is unacceptable in the adult language. However, some unrelated languages do seem to allow for doubling of the head.28 Example (47a) is from Kuno (1973, p. 237); (47b) is from Cinque (2011, (24b)).
Such data are potentially problematic for certain relativization theories, including the raising analysis to be discussed in section 4.1. For each case to be valid, it must be made plausible that the construction is neither correlative nor appositive (recall (38b) concerning ARCs).
3.8 Free Relatives
Free relatives are relative constructions without a visible head noun or external determiner. They occur in languages with all major relativization types, as is illustrated in (48a-d), cited from Lehmann (1984, pp. 97, 295, 295, and 124), respectively.
Typically, free relatives are nominalized clauses and occur in noun phrase positions. The semantics is maximalizing, which implies that only a definite or universal interpretation is possible, and that stacking is excluded; see (49a), from Grosu (2002, p. 148).29 Note that in languages with relative pronouns, the internal head position of a FR may sometimes be expressed overtly as a restriction on the operator, as in (49b):
Theoretical discussion in syntax concerns the status of an external determiner and—possibly—head position, and the surface position of the relative pronoun: is it RC-internal (in the CP domain) or directly related to the external determiner position, which would otherwise be abstract?30
It may be relevant that in addition to headed RCs and FRs, there are lightheaded or semi-FRs, in which there is just an external determiner, for example an article, demonstrative, or quantifier, but no external head noun.31 It is not easy to decide what counts as lightheaded, since pronominal forms and strong paradigms indicating noun ellipsis are frequently used. Notably, these pattern with regular headed relatives in relevant respects: they can be indefinite, as in (50), and they do not display (Case) matching effects.
It appears that FRs can also be nonnominal, especially when they are adverbial:32
Special mention should be given to so-called transparent free relatives, which involve a phrase—italicized in (52)—that is predicated of the wh-operator and that intuitively functions as the semantic and syntactic nucleus of the construction.
It is debated to what extent these constructions differ from regular free relatives, and how they are to be analyzed.33
3.9 Non-Finite Relative Clauses and Modal Existential Wh-Constructions
Relative clauses can be non-finite. Example (53a) illustrates an infinitival relative in English;34 (53b) and (53c) may be considered to contain a prenominal present and past participial relative clause, respectively. However, the distinction between (deverbal) adjectives and reduced forms of relative clauses is unclear, if it exists at all.
The construction in (53c) appears to be a hidden passive, cf. the clothes washed by my father. Participial relatives, then, always involve subject relativization in English and related languages. But this is not universally the case, witness examples like the following, from Lehmann (1984, p. 50):
So-called modal existential wh-constructions (or ‘irrealis free relatives’), are illustrated in (55), from (Šimík, 2011, pp. 2 and 6):
Here, an infinitival or subjunctive wh-relative clause is embedded under an existential predicate, and receives a modal interpretation.35
3.10 Cleft and Pseudocleft Constructions
Cleft and pseudocleft constructions are predicational focus constructions that share certain characteristics with appositive and restrictive relative clauses, depending on the subtype. English it-clefts are illustrated in (56). As (56b) shows, the focus does not need to be a noun phrase.
Pseudoclefts contain a free relative clause; see (57) in English. The interpretation can be either specificational or predicational, that is, the predicate either contributes to the identification of the referent of the subject or attributes a property to it.
There is a substantial literature on the structure and meaning of clefts and pseudoclefts, which cannot be treated here.36
3.11 Pseudo-Relatives and Quasi/V2-Relatives
Pseudo-relatives are constructions that look like subject relatives, documented for Romance languages (and Greek) in particular; see (58), from Moulton and Grillo (2015, p. 2):
Given their interpretation, possible positions, and certain restrictions (including tense agreement and the form of the relative linker), they have been analyzed differently from regular relative clauses.37
Quasi-relatives or V2-relatives are found in Dutch and German. In these languages, main clauses are verb second, but subordinate clauses are (normally) not. Therefore, the V2 order in the apparent relative clause in (59) is remarkable:
There is consensus in the literature, however, that such examples do not involve relativization at all, for various reasons: the linker is a demonstrative, not a relative pronoun; the relative clause is sentence-final; the coordinator en ‘and’ can be inserted. Therefore, despite the cohesive intonation pattern, the connection between the clauses must involve parataxis of some sort.38
4. Further Theoretical Discussion
Two general theoretical questions have received ample attention in the literature: the connectivity problem and the modification problem. (i) How can the pivotal role of the relative head be explained, and specifically, how is the variable corresponding to the relative head represented internally to the relative clause? (ii) How and at which hierarchical level is the relative clause integrated in its syntactic context (potentially depending on the type of relative clause)? These issues are briefly addressed in the two following subsections.
4.1 Raising and Matching
In an externally headed relative construction containing a relative pronoun, it seems evident that a pronominal representation of the head is fronted within the relative clause. If there is only a complementizer or no clausal linker at all, an abstract operator can be postulated; recall the discussion in section 3.1:
For decades, variants of (60) have been the most common analysis of (head-external) relative constructions.39
However, there are reasons to assume a full representation of the head within the relative clause; this has to do with reconstruction and other connectivity effects.40 In (61), for example, the head must be reconstructed into the relative clause in order to obtain the correct interpretation. The relevant thematic position is indicated with an underscore. In (61a), take advantage (of) is a fixed expression (idiomatic collocation); in (61b) the anaphor each other is bound by the RC-internal subject John and Bill; in (61c) the pronominal variable her is bound by (hence its interpretation covaries with) the RC-internal quantified expression every granny; in (61d) there is an inverse scope interpretation every > six, that is, ‘for every insect: it has six legs’ rather than the nonsensical interpretation ‘for six particular legs: every insect has them’, which would correspond to the surface order.
A possible solution is the assumption of head raising (or promotion): the head NP finds its origin at the thematic base position inside the relative clause, and is subsequently raised to its surface position. An important advantage of this idea is that it allows a generalization over head-external and head-internal relative constructions (cf. sections 3.6–3.8).
Potential (technical) difficulties involved concern the status of relative pronouns, the distribution of case (cf. (20a)), and restrictions on movement. A familiar implementation of raising is sketched in (62). Here, the relative clause is analyzed as the complement of the external determiner (hence the entire construction comes down to a nominalized clause, which is evident for (semi-)free relatives and IHRCs).41 A relative pronoun can be viewed as a determiner related to the head NP.42 The DPrel is generated RC-internally in its thematic base position, and subsequently Aʹ‑moved to the CP-domain (note that Drel, whether overt or abstract, is an operator). Finally, the head NP is moved to its surface position left of the relative pronoun, where it can be case-associated with the external determiner.
It is important to note that this straightforwardly solves the pertinent coindexing problem mentioned around (24) of section 3.1.
Despite the possibility of reconstruction in sentences like (61), there are also non- or even anti-reconstruction effects. For instance, in (63a) there is no idiom reconstruction; in the most sensible interpretation of (63b), two scopes over every; in (63c), there is no Principle C effect, which says that a pronoun like he cannot bind a referential expression like John (which would be the case after reconstruction); in (63d), the anaphor can be bound by the subject of the matrix clause.
An alternative analysis is the matching analysis43 shown in (64), which can be considered an enhanced version of (60). There are both an external and an internal representation of the head NP. The internal one is obligatorily deleted by means of NP-ellipsis (which raises questions to be answered).
Originally, it was assumed that only the external interpretation of NP is compatible with this structure (as in the regular head-external analysis). Relatedly, it could be said that raising and matching do not exclude each other but coexist, giving rise to the interpretations in (61) as well as (63).44 It has also been claimed, however, that the raising analysis is superfluous if the internal representation of the head in (64) can also be interpreted, that is, if matching can account for reconstruction as well as non-reconstruction.45 The most intricate (and debated) examples that may support such an approach are sentences with conflicting requirements, that is, where both reconstruction and non-reconstruction are necessary for the interpretation:
Here the interpretation of the collocation take a picture requires reconstruction of the relative head picture of John/himself into the complement position of took inside the relative clause, but in (65a) this would lead to a principle C violation inside the relative clause (he would then bind John), and in (65b) the anaphor himself could then not be bound by the nonlocal antecedent Paul across the finite clause boundary.46
4.2 Level of Attachment and the Syntactic Status of Appositives
A restrictive relative clause must be attached within the scope of an external determiner. In a traditional head-external or matching analysis, it could be an adjunct to the external NP or a complement of the head N (see section 3.1). In many current raising analyses, the RC is directly the complement of the external D, as in (62).
For appositive (non-restrictive) relatives, the situation is different. First, as discussed in section 3.2, these relatives outscope external determiners (in fact, by definition), with structural consequences. However, it has been established that it is insufficient to simply adjoin them at the DP level. While they are formally embedded clauses, appositives do bear (semantic) similarities to main clauses: they embody speech acts, carry illocutionary force, and may contain discourse particles and speaker-oriented adverbs:47
Furthermore, they are outside the scope of quantifiers higher up in the matrix. Compare the appositive construction in (67a) to the restrictive one in (67b):
It is also relevant to note that none of the reconstruction data summarized in (61) in the previous subsection carry over to appositives; see (68), for instance.
Thus, appositives are much like parentheticals, and it is tempting to treat them as orphans, that is, not part of sentential syntax at all.48 However, there are also arguments to incorporate them in syntax (but in a special way).49 For instance, it is clear that they do have a fixed position adjacent to the antecedent (modulo right-extraposition to the edge of the containing clause), contrary to independent parentheticals.
Oxford Bibliographies Online
Oxford Bibliography on Relative Clauses. doi:10.1093/obo/9780199772810-0120
There is a large and ever-growing amount of published work on relative constructions. This section contains a few useful points of access to the existing literature that largely complement specifically directed references in sections 1–4, emphasizing recent overview literature and state-of-the-art work rather than classics. A much more elaborate annotated list of studies is the Oxford Bibliography on Relative Clauses.
A general, theoretically oriented overview of headed relatives, including a bibliography, is Bianchi (2002); it is complemented by Grosu (2002), which discusses “peripheral” types of RCs. A more typologically oriented impression is Andrews (2007). See also de Vries (2002) and the paper collection in Alexiadou et al. (2000) on various theoretical and typological aspects of RCs. A recent accessible theoretical overview can be found in Salzmann (2017).
A rich source of cross-linguistic data is Lehmann (1984). More recently, one can also consult the WALS Online database (Dryer & Haspelmath, 2011; see especially chapters 60, 90, 96, 122, 123). Smits (1988) contains more details about the Germanic and Romance languages; Murelli (2011) focuses on non-standard varieties from different language families in Europe; the contributions in Comrie and Estrada-Fernández (2012) address languages in the Americas. A relatively recent update on RC-related universals and tendencies is de Vries (2005). Hendery (2012) discusses the typology of RCs from a diachronic perspective.
Wu (2011) gathers findings about prenominal relatives. Studies on correlatives are collected in Lipták (2009), which also contains an introduction to the topic. A division of internally headed relatives into subtypes is discussed in Grosu (2012). An overview on Free Relatives is Van Riemsdijk (2006); a more recent discussion, especially on Transparent Free Relatives, is Grosu (2016). Amount relatives are discussed in detail in Grosu and Landman (2017). Concerning modal existential wh-constructions, see especially Šimík (2011). Quasi/V2-relatives are examined in de Vries (2012). A recent theoretical discussion of appositive relative clauses can be found in Griffiths (2015); discourse functions of nonrestrictives are investigated in Loock (2010).
Alexiadou, A., Law, P., Meinunger, A., & Wilder, C. (Eds.). (2000). The syntax of relative clauses. Amsterdam: John Benjamins.Find this resource:
Andrews, A. (2007). Relative clauses. In T. Shopen (Ed.), Language typology and syntactic description (2nd ed., Vol. 2, pp. 206–236). Cambridge, UK: Cambridge University Press.Find this resource:
Aoun, J., & Li, Y. (2003). Essays on the representational and derivational nature of grammar: The diversity of wh-constructions. Cambridge, MA: MIT Press.Find this resource:
Arnold, D. (2007). Non-restrictive relatives are not orphans. Journal of Linguistics, 43, 272–309.Find this resource:
Basilico, D. (1996). Head position and internally headed relative clauses. Language, 72, 498–533.Find this resource:
Bertollo, S. (2014). On relatives with a null head: German free relative clauses and clefts (Doctoral Dissertation). University of Padova.Find this resource:
Bhatt, R. (2002). The raising analysis of relative clauses: Evidence from adjectival modification. Natural Language Semantics, 10, 43–90.Find this resource:
Bhatt, R. (2003). Locality in correlatives. Natural Language & Linguistic Theory, 210, 485–541.Find this resource:
Bianchi, V. (1999). Consequences of antisymmetry: Headed relative clauses. Berlin: Mouton de Gruyter.Find this resource:
Bianchi, V. (2002). Headed relative clauses in generative syntax, Part I and Part 2. Glot International, 6(7), 197–204, 6(8), 235–247.Find this resource:
Bodomo, A., & Hiraiwa, K. (2010). Relativization in Dàgáárè and its typological implications: Left-headed but internally-headed. Lingua, 120(4), 953–983.Find this resource:
Boef, E. (2013). Doubling in relative clauses: Aspects of morphosyntactic microvariation in Dutch (Doctoral Dissertation). Utrecht University. LOT Dissertation Series 317.Find this resource:
Bonneau, J. (1992). The structure of internally headed relative clauses: Implications for configurationality (Doctoral Dissertation). McGill University, Montreal.Find this resource:
Bresnan, J., & Grimshaw, J. (1978). The syntax of free relatives in English. Linguistic Inquiry, 9, 331–391.Find this resource:
Busch, J. L., & Schumann, F. (2016). Unspecific indefinites and (non-)restrictive relative clauses. Lingua, 181, 1–35.Find this resource:
Caponigro, I. (2003). Free not to ask: On the semantics of free relatives and wh-words cross-linguistically (Unpublished doctoral dissertation). University of California, Los Angeles.Find this resource:
Cardoso, A. (2017). Portuguese relative clauses in synchrony and diachrony. Oxford: Oxford University Press.Find this resource:
Cardoso, A., & de Vries, M. (2010). Internal and external heads in appositive constructions (Unpublished manuscript). University of Lisbon and University of Groningen.
Carlson, G. (1977). Amount relatives. Language, 53, 520–542.Find this resource:
Catasso, N. (2014). Wie viele Jungs haben Anna geküsst? Zum besonderen Status von V2-Relativsätzen im gesprochenen Deutsch. Linguistik online, 67(5), 45–68.Find this resource:
Chomsky, N. (1965). Aspects of the theory of syntax. Cambridge, MA: MIT Press.Find this resource:
Chomsky, N. (1977). On wh-movement. In P. Culicover, T. Wasow, & A. Akmajian (Eds.), Formal syntax (pp. 71–132). New York: Academic Press.Find this resource:
Cinque, G. (1981). On Keenan and Comrie’s primary relativization constraint. Linguistic Inquiry, 12, 293–308.Find this resource:
Cinque, G. (2008). Two types of non-restrictive relatives. Empirical Issues in Syntax and Semantics, 7, 99–137.Find this resource:
Cinque, G. (2009). Five notes on correlatives. University of Venice Working Papers in Linguistics, 19, 35–60.Find this resource:
Cinque, G. (2010). On a selective “violation” of the Complex NP Constraint. In J.-W. Zwart & M. de Vries (Eds.), Structure preserved: Studies in syntax for Jan Koster (pp. 81–89). Amsterdam: John Benjamins.Find this resource:
Cinque, G. (2011). On double-headed relative clauses. Linguística. Revista de Estudos Linguísticos da Universidade do Porto, 1(6), 67–91.Find this resource:
Cinque, G. (2015). Three phenomena discriminating between “raising” and “matching” relative clauses. Semantics-Syntax Interface, 2(1), 1–27.Find this resource:
Citko, B. (2001). Deletion under identity in relative clauses. Proceedings of the North East Linguistic Society (NELS), 31, 131–145.Find this resource:
Citko, B. (2002). (Anti)reconstruction effects in free relatives: A new argument against the COMP account. Linguistic Inquiry, 33, 507–511.Find this resource:
Citko, B. (2004). On headed, headless and light-headed relatives. Natural Language & Linguistic Theory, 22, 95–126.Find this resource:
Cole, P. (1987). The structure of internally headed relative clauses. Natural Language & Linguistic Theory, 5(2), 277–302.Find this resource:
Comrie, B., & Estrada-Fernández, Z. (Eds.). (2012). Relative clauses in languages of the Americas: A typological overview. Amsterdam: John Benjamins.Find this resource:
Constant, N. (2011). Re-diagnosing appositivity: Evidence for prenominal appositives from Mandarin. Proceedings of Chicago Linguistics Society (CLS), 47.Find this resource:
Culy, C. (1990). The syntax and semantics of internally headed relative clauses (Unpublished doctoral dissertation). Stanford University.Find this resource:
Dehé, N. (2017). Parentheticals in spoken English: The syntax-prosody relation. Cambridge, UK: Cambridge University Press.Find this resource:
Del Gobbo, F. (2007). On the syntax and semantics of appositive relative clauses. In N. Dehé & Y. Kavalova (Eds.), Parentheticals (pp. 173–201). Amsterdam: John Benjamins,Find this resource:
Del Gobbo, F. (2010). On Chinese appositive relative clauses. Journal of East Asian Linguistics, 19, 385–417.Find this resource:
Demirdache, H. (1991). Resumptive chains in restrictive relatives, appositives and dislocation structures (Doctoral Dissertation). MIT.Find this resource:
Doherty, C. (2000). Clauses without “that”: The case for bare sentential complementation in English. London: Routledge.Find this resource:
Donati, C., & Cecchetto, C. (2011). Relabeling heads: A unified account for relativization structures. Linguistic Inquiry, 42, 519–560.Find this resource:
Douglas, J. (2016). The syntactic structures of relativisation (Doctoral Dissertation). University of Cambridge.Find this resource:
Dryer, M., & Haspelmath, M. (Eds.). (2011). The world atlas of language structures online. Munich: Max Planck Digital Library.Find this resource:
Emonds, J. (1979). Appositive relatives have no properties. Linguistic Inquiry, 10, 211–243.Find this resource:
Fabb, N. (1990). The difference between English restrictive and nonrestrictive relative clauses. Journal of Linguistics, 26, 57–78.Find this resource:
Gärtner, H.-M. (2001). Are there V2 relative clauses in German? Journal of Comparative Germanic Linguistics, 3, 97–141.Find this resource:
Givón, T. (1984). Syntax, a functional-typological approach. Amsterdam: John Benjamins.Find this resource:
Göbbel, E. (2013). Extraposition of relative clauses: Phonological solutions. Lingua, 136, 77–102.Find this resource:
Gondra, A. (2016). Head external analysis, head raising analysis or matching analysis? Let’s ask experimental syntax. Lingua, 179, 57–75.Find this resource:
Gračanin-Yuksek, M. (2008). Free relatives in Croatian: Arguments for the COMP account. Linguistic Inquiry, 39, 275–294.Find this resource:
Griffiths, J. (2015). On appositives (Doctoral Dissertation). University of Groningen. LOT Dissertation Series 389.Find this resource:
Griffiths, J., & de Vries, M. (2013). The syntactic integration of appositives: Evidence from fragments and ellipsis. Linguistic Inquiry, 44, 332–344.Find this resource:
Griffiths, J., & de Vries, M. (to appear). Parenthesis: Syntactic integration or orphanage? Linguistic Inquiry.Find this resource:
Groos, A., & van Riemsdijk, H. (1981). Matching effects in free relatives: A parameter of core grammar. In A. Belletti, L. Brandi, & L. Rizzi (Eds.), Theory of markedness in generative grammar: Proceedings of the 1979 GLOW Conference (pp. 171–216). Pisa: Scuola Normale Superiore.Find this resource:
Grosu, A. (1996). The proper analysis of “missing-P” free relative constructions. Linguistic Inquiry, 27, 257–293.Find this resource:
Grosu, A. (2002). Strange relatives at the interface of two millennia. Glot International, 6(6), 145–167.Find this resource:
Grosu, A. (2003). A unified theory of ‘standard’ and ‘transparent’ free relatives. Natural Language & Linguistic Theory, 21, 247–331.Find this resource:
Grosu, A. (2012). Towards a more articulated typology of internally-headed relative constructions: The semantic connection. Language and Linguistics Compass, 6, 447–476.Find this resource:
Grosu, A. (2016). The semantics, syntax, and morphology of Transparent Free Relatives revisited: A comparison of two approaches. Natural Language & Linguistic Theory, 34, 1245–1280.Find this resource:
Grosu, A., & Landman, F. (1998). Strange relatives of the third kind. Natural Language Semantics, 6(2), 125–170.Find this resource:
Grosu, A., & Landman, F. (2017). Amount relatives. In M. Everaert & H. van Riemsdijk (Eds.), The Wiley Blackwell Companion to Syntax (2nd ed.). Malden, MA: Blackwell.Find this resource:
Güneş, G. (2015). Deriving prosodic structures (Doctoral Dissertation). University of Groningen. LOT Dissertation Series 396.Find this resource:
Hartmann, K., & Veenstra, T. (Eds.). (2013). Cleft structures. Amsterdam: John Benjamins.Find this resource:
Heck, F. (2008). On pied piping: Wh-movement and beyond. Berlin: Mouton de Gruyter.Find this resource:
Hendery, R. (2012). Relative clauses in time and space. Amsterdam: John Benjamins.Find this resource:
Hiraiwa, K. (2005). Dimensions of symmetry in syntax: Agreement and clausal architecture (Unpublished Doctoral Dissertation). Massachusetts Institute of Technology.Find this resource:
Hoeksema, J. (1986). An account of relative clauses with split antecedents. Proceedings of the West Coast Conference on Formal Linguistics (WCCFL), 5, 68–86.Find this resource:
Holler, A. (2005). Weiterführende Relativsätze: Empirische und theoretische Aspekte. Berlin: Akademie Verlag.Find this resource:
Hoshi, K. (1995). Structural and interpretive aspects of head-internal and head-external relative clauses (Unpublished Doctoral Dissertation). University of Rochester.Find this resource:
Hulsey, S., & Sauerland, U. (2006). Sorting out relative clauses. Natural Language Semantics, 14, 111–137.Find this resource:
Itô, J. (1986). Head movement at LF and PF: The syntax of head-internal relatives in Japanese. University of Massachusetts Occasional Papers in Linguistics, 11, 109–138.Find this resource:
Jackendoff, R. (1977). X´-syntax: A study of phrase structure. Cambridge, MA: MIT Press.Find this resource:
Jacobson, P. (1995). On the quantificational force of English free relatives. In E. Bach et al. (Eds.), Quantification in natural languages (pp. 451–486). Dordrecht: Kluwer.Find this resource:
Kayne, R. (1994). The antisymmetry of syntax. Cambridge, MA: MIT Press.Find this resource:
Kayne, R. (2015). A note on some even more unusual relative clauses (Unpublished manuscript). New York University.Find this resource:
Keenan, E., & Comrie, B. (1977). Noun phrase accessibility and universal grammar. Linguistic Inquiry, 8, 63–99.Find this resource:
Keenan, E., & Comrie, B. (1979). Noun phrase accessibility revisited. Language, 55, 649–664.Find this resource:
Kidd, E. (Ed.). (2011). The acquisition of relative clauses: Processing, typology and function. Amsterdam: John Benjamins.Find this resource:
Kim, M.-J. (2007). Formal linking in internally headed relatives. Natural Language Semantics, 15(4), 279–315.Find this resource:
Kluck, M. (2011). Sentence amalgamation (Doctoral Dissertation). University of Groningen. LOT Dissertation Series 285.Find this resource:
Kuno, S. (1973). The structure of the Japanese language. Cambridge, MA: MIT Press.Find this resource:
Larson, R. (1987). ‘Missing prepositions’ and the analysis of English free relative clauses. Linguistic Inquiry, 18, 239–266.Find this resource:
Larson, R., & Takahashi, N. (2007). Order and interpretation in prenominal relative clauses. In M. Kelepir & B. Öztürk (Eds.), MIT Working Papers in Linguistics 54: Proceedings of the Workshop on Altaic Formal Linguistics II (pp. 101–120). Cambridge, MA: MITWPL.Find this resource:
Lehmann, C. (1984). Der Relativsatz. Tübingen: Gunter Narr.Find this resource:
Lin, J.-W., & Tsai, W.-T. (2015). Restricting non-restrictive relatives in Mandarin Chinese. In A. Li, A. Simpson, & W.-T. Tsai (Eds.), Chinese syntax in a cross-linguistic perspective (pp. 100–127). Oxford: Oxford University Press.Find this resource:
Lipták, A. (Ed.). (2009). Correlatives cross-linguistically. Amsterdam: John Benjamins.Find this resource:
Loock, R. (2010). Appositive relative clauses in English: Discourse functions and competing structures. Amsterdam: John Benjamins.Find this resource:
McCloskey, J. (2006). Resumption. In M. Everaert & H. van Riemsdijk (Eds.), The Wiley Blackwell Companion to Syntax (pp. 94–117). Oxford: Blackwell.Find this resource:
McKinney-Bock, K. (2013). Deriving split-antecedent relative clauses. University of Pennsylvania Working Papers in Linguistics, 19, 112–122.Find this resource:
Moulton, K., & Grillo, N. (2015). Sorting out pseudo-relatives: Clausal determiners and mediated AGREE. Unpublished Manuscript.
Murelli, A. (2011). Relative constructions in European non-standard varieties. Berlin: De Gruyter/Mouton.Find this resource:
Ojea, A. (2013). A uniform account of headless relatives in Spanish. Language Sciences, 40, 200–211.Find this resource:
Ott, D. (2011). A note on free relative clauses in the theory of phases. Linguistic Inquiry, 42, 183–192.Find this resource:
Ott, D. (2016). Fragment anchors do not support the syntactic integration of appositive relatives: Reply to Griffiths & de Vries 2013. Linguistic Inquiry, 47, 580–590.Find this resource:
Pankau, A. (2015). The matching analysis of relative clauses: Evidence from Upper Sorbian. In Y. Oseki et al. (Eds), Proceedings of the 24th meeting of Formal Approaches to Slavic Linguistics. Ann Arbor: Michigan Slavic Publications.Find this resource:
Pérez-Leroux, A. T. (1995). Resumptives in the acquisition of relative clauses. Language Acquisition, 4, 105–138.Find this resource:
Perlmutter, D., & Ross, J. (1970). Relative clauses with split antecedents. Linguistic Inquiry, 1, 365.Find this resource:
Rafel, J. (2000). Complex small clauses (Unpublished Doctoral Dissertation). University of Barcelona.Find this resource:
Rebuschi, G. (2003). Basque semi-free relative clauses and the structure of DPs. Lapurdum, 8, 457–477.Find this resource:
Reeve, M. (2012). Clefts and their relatives. Amsterdam: John Benjamins.Find this resource:
van Riemsdijk, H. (2000). Free relatives inside out: Transparent free relatives as grafts. In B. Rozwadowska (Ed.), Proceedings of the 1999 PASE Conference (pp. 223–233). Wrocław: Aksel.Find this resource:
van Riemsdijk, H. (2006). Free relatives. In M. Everaert & H. van Riemsdijk (Eds.), The Wiley Blackwell Companion to Syntax (Vol. 2, pp. 338–382). Oxford: Blackwell.Find this resource:
Ross, J. (1967). Constraints on variables in syntax (Unpublished Doctoral Dissertation). MIT.Find this resource:
Safir, K. (1986). Relative clauses in a theory of binding and levels. Linguistic Inquiry, 17, 663–689.Find this resource:
Salzmann, M. (2017). Reconstruction and resumption in indirect Aʹ-dependencies. Berlin: De Gruyter/Mouton.Find this resource:
Schachter, P. (1973). Focus and relativization. Language, 49, 19–46.Find this resource:
Sells, P. (1984). Syntax and semantics of resumptive pronouns (Unpublished Doctoral Dissertation). University of Massachusetts, Amherst.Find this resource:
Sells, P. (1985). Restrictive and non-restrictive modification. CSLI Report, 27–28, 1–33.Find this resource:
Shimoyama, J. (1999). Internally headed relative clauses in Japanese and E-type anaphora. Journal of East Asian Linguistics, 8(2), 147–182.Find this resource:
Sichel, I. (2014). Resumptive pronouns and competition. Linguistic Inquiry, 45, 655–693.Find this resource:
Šimík, R. (2011). Modal existential wh-constructions (Doctoral Dissertation). University of Groningen. LOT Dissertation Series 269.Find this resource:
Smith, C. (1964). Determiners and relative clauses in a generative grammar of English. Language, 40, 37–52.Find this resource:
Smits, R. (1988). The relative and cleft constructions of the Germanic and Romance languages. Dordrecht: Foris.Find this resource:
Srivastav, V. (1991). The syntax and semantics of correlatives. Natural Language & Linguistic Theory, 9(4), 637–686.Find this resource:
Stockwell, R., Schachter, P., & Partee, B. (1973). The major syntactic structures of English. New York: Holt.Find this resource:
Taraldsen, T. (1981). The theoretical interpretation of a class of marked extractions. In A. Belletti, L. Brandi, & L. Rizzi (Eds.), Theory of markedness in generative grammar (pp. 475–516). Pisa: Scuola normale superiore di Pisa.Find this resource:
Truckenbrodt, H. (2015). Intonation phrases and speech acts. In M. Kluck, D. Ott, & M. de Vries (Eds.), Parenthesis and ellipsis: Cross-linguistic and theoretical perspectives (pp. 301–349). Berlin: De Gruyter/Mouton.Find this resource:
Vercauteren, A. (2016). A conspiracy theory for clefts: The syntax and interpretation of cleft constructions (Unpublished Doctoral Dissertation). University Nova of Lisbon and Ghent University.Find this resource:
Vergnaud, J.-R. (1974). French relative clauses (Unpublished Doctoral Dissertation). MIT.Find this resource:
de Vries, M. (2002). The syntax of relativization (Doctoral Dissertation). University of Amsterdam. LOT Dissertation Series 53.Find this resource:
de Vries, M. (2005). The fall and rise of universals on relativization. Journal of Universal Language, 6, 125–157.Find this resource:
de Vries, M. (2006a). Possessive relatives and (heavy) pied piping. Journal of Comparative Germanic Linguistics, 9, 1–52.Find this resource:
de Vries, M. (2006b). The syntax of appositive relativization: On specifying coordination, false free relatives and promotion. Linguistic Inquiry, 37, 229–270.Find this resource:
de Vries, M. (2012). Parenthetical main clauses—or not? On appositives and quasi-relatives. In L. Aelbrecht, L. Haegeman, & R. Nye (Eds.), Main clause phenomena: New horizons (pp. 177–201). Amsterdam: John Benjamins.Find this resource:
de Vries, M. (2017). Across-the-board phenomena. In M. Everaert & H. van Riemsdijk (Eds.), The Wiley Blackwell Companion to Syntax (2nd ed., pp. 1–31). Malden, MA: Blackwell. doi:10.1002/9781118358733.wbsyncom094Find this resource:
Watanabe, A. (2004). Parametrization of quantificational determiners and head-internal relatives. Language and Linguistics, 5, 59–97.Find this resource:
Wilder, C. (1999). Transparent free relatives. Proceedings of the West Coast Conference on Formal Linguistics (WCCFL), 17, 685–699.Find this resource:
Williamson, J. (1987). An indefinite restriction for relative clauses in Lakhota. In E. Reuland & A. ter Meulen (Eds.), The Representation of (In)definiteness (pp. 168–190). Cambridge, MA: MIT Press.Find this resource:
Wu, T. (2011). The syntax of prenominal relative clauses: A typological study. Linguistic Typology, 15, 569–623.Find this resource:
Zhang, N. N. (2007). The syntactic derivations of split antecedent relative clause constructions. Taiwan Journal of Linguistics, 5, 19–48.Find this resource:
Zwart, J.-W. (2000). A head raising analysis of relative clauses in Dutch. In A. Alexiadou et al. (Eds.), The syntax of relative clauses (pp. 349–385). Amsterdam: John Benjamins.Find this resource:
Zwart, J.-W. (2005). Iets over zgn. V2-relatieven in het Nederlands. Nederlandse taalkunde, 10, 59–81.Find this resource:
(2.) See for example contributions in Kidd (2011) and references there for discussion. Note that there is a substantial and ever-growing literature about processing, parsing, and L1/2 acquisition of RCs. These topics fall outside the scope of this article.
(9.) The traditional view was still present in Chomsky (1965) and Ross (1967). The issue concerning scope was formulated in Stockwell, Schachter, and Partee (1973), among others. Jackendoff (1977) proposed a more articulate hierarchical structure, which paved the way for modern approaches.
(11.) See Taraldsen (1981) for an exception in Swedish, and Cinque (2010) for a brief overview of problematic data from other languages and further references. The use of an intrusive resumptive pronoun has a reported alleviating effect in English:
((i)) ? the man who I heard the claim that Peter knows him
Such an intrusive pattern is not universally allowed, however. In languages where resumptive pronouns are a regular relative strategy (e.g., Hebrew, Irish), the use of intrusive resumptives can create the appearance of an across-the-board insensitivity to island boundaries in RCs. However, bound reading tests can be used to distinguish both configurations: intrusive pronouns are never bound variables, and island-insensitivity is an illusion. See Sells (1984), McCloskey (2006), Sichel (2014) for discussion and references.
(12.) The alleged difference between appositive and restrictive RCs with respect to pied piping (going back to Ross, 1967) is contradicted in De Vries (2006a): RRCs do allow for (very) heavy pied piping, provided they are introduced by a preposition. See also Heck (2008) for relevant discussion and references.
(24.) For discussion see Grosu (2012), and earlier work by Itô (1986), Williamson (1987), Cole (1987), Culy (1990), Bonneau (1992), Hoshi (1995), Shimoyama (1999), Watanabe (2004), Hiraiwa (2005), and Kim (2007), among others.
(25.) See Srivastav (1991) and Grosu (2002). Surprisingly, Cinque (2009) also reports the existence of non-restrictive correlatives in Marathi and a few other languages, suggesting, however, that these could be reanalyzed as postnominal ARCs with a deleted antecedent.
(27.) See de Vries (2017) for some discussion and further references. Furthermore, Kayne (2015) reports that even true instances of multiple relativization can be found in English; see (i) and (ii). It is to be noted, however, that these involve appositive constructions.
(28.) See Cinque (2011) for an overview and discussion. Cinque’s term “double-headed RCs” is not to be confused with the qualitatively different constructions involving multiple relativization discussed in section 3.6.
(30.) For discussion and some opposing views, see Bresnan and Grimshaw (1978), Groos and Van Riemsdijk (1981), Citko (2002), De Vries (2002), Gračanin-Yuksek (2008), Ott (2011), Ojea (2013), and Bertollo (2014).
(40.) See early work by Schachter (1973), Vergnaud (1974), Carlson (1977), and later Kayne (1994), Bianchi (1999), De Vries (2002), and others. Cardoso (2017) adds an interesting argument for raising that involves a stranded modifier inside the relative clause. Furthermore, see Zwart (2000) concerning selection effects between a relative pronoun and a quantifier head noun; Bhatt (2002) concerning scope effects with certain adjectives; Pankau (2015) concerning relativization in antipronominal contexts; Gondra (2016) for experimental confirmation of reconstructed readings; and Grosu and Landman (2017) concerning amount relatives.
(46.) See Douglas (2016) and Salzmann (2017) for recent critical discussions of all the complex (non-)reconstruction data. Note that additional issues to take into account are the possibility of vehicle change and late merger of modifiers.