Show Summary Details

Page of

Printed from Oxford Research Encyclopedias, Linguistics. Under the terms of the licence agreement, an individual user may print out a single article for personal use (for details see Privacy Policy and Legal Notice).

date: 06 July 2022

Coordination in Syntaxfree

Coordination in Syntaxfree

  • Grant GoodallGrant GoodallDepartment of Linguistics, University of California, San Diego


The term coordination refers to the juxtaposition of two or more conjuncts often linked by a conjunction such as and or or. The conjuncts (e.g., our friend and your teacher in Our friend and your teacher sent greetings) may be words or phrases of any type. They are a defining property of coordination, while the presence or absence of a conjunction depends on the specifics of the particular language. As a general phenomenon, coordination differs from subordination in that the conjuncts are typically symmetric in many ways: they often belong to like syntactic categories, and if nominal, each carries the same case. Additionally, if there is extraction, this must typically be out of all conjuncts in parallel, a phenomenon known as Across-the-Board extraction. Extraction of a single conjunct, or out of a single conjunct, is prohibited by the Coordinate Structure Constraint. Despite this overall symmetry, coordination does sometimes behave in an asymmetric fashion. Under certain circumstances, the conjuncts may be of unlike categories or extraction may occur out of one conjunct, but not another, thus yielding apparent violations of the Coordinate Structure Constraint. In addition, case and agreement show a wide range of complex and sometimes asymmetric behavior cross-linguistically. This tension between the symmetric and asymmetric properties of coordination is one of the reasons that coordination has remained an interesting analytical puzzle for many decades.

Within the general area of coordination, a number of specific sentence types have generated much interest. One is Gapping, in which two sentences are conjoined, but material (often the verb) is missing from the middle of the second conjunct, as in Mary ate beans and John _ potatoes. Another is Right Node Raising, in which shared material from the right edge of sentential conjuncts is placed in the right periphery of the entire sentence, as in The chefs prepared __ and the customers ate __ [a very elaborately constructed dessert]. Finally, some languages have a phenomenon known as comitative coordination, in which a verb has two arguments, one morphologically plural and the other comitative (e.g., with the preposition with), but the plural argument may be understood as singular. English does not have this phenomenon, but if it did, a sentence like We went to the movies with John could be understood as John and I went to the movies.


  • Semantics
  • Syntax

Coordination is a syntactic phenomenon in which two or more elements, known as conjuncts, are linked together, often with a conjunction (also known traditionally as a coordinating conjunction). In (1), for instance, the conjuncts are John and Mary, and the conjunction is and.


In general, conjuncts may be of any category and may be either individual words, as in (2), or phrases, as in (3).



Conjunctions are typically either and or or. Coordinate structures may also occur through simple juxtaposition of the conjuncts. In English, this most often occurs when there are three or more conjuncts, as in (4), where the first two conjuncts are simply juxtaposed.


In many languages, this is common even with only two conjuncts, as in the Mandarin Chinese example in (5).


The conjuncts, on the other hand, are a defining component of coordinate structures. As the term coordination suggests, the conjuncts are typically in a symmetric relationship to each other, as opposed to the asymmetric relationship typical of subordination. This symmetry may be seen in a number of ways. First, each conjunct is equally able to stand on its own in place of the larger conjoined structure. Given (3a), for example, the two sentences in (6) are also possible.


Second, each conjunct typically has the same case, the case that the larger conjoined structure would have. In (3a), this structure is in a nominative position, and each conjunct is nominative as well, as may be seen when pronouns are used, as in (7).


Third, each conjunct is typically of the same syntactic category. That is, even when each conjunct would be well formed on its own in the sentence, noticeable degradation occurs when the two conjuncts do not match in category, as seen in (8).


This phenomenon is known as the Law of the Coordination of Likes. Finally, there is symmetry between the conjuncts in the sense that when there is extraction out of one, there typically needs to be extraction out of both, as in (9).


Although examples such as the above, in which the conjuncts behave symmetrically, are widespread and often taken to be typical of coordination, many types of asymmetric behavior do also occur. For example, although it is generally true that each conjunct is able to stand on its own in place of the larger conjoined structure, this is not always true. In (10), for instance, the second conjunct is not ordinarily a possible object for on, but it is possible when conjoined with something else.


Similarly, the two conjuncts typically match in case, having the case that would be appropriate for the larger conjoined structure, but this expectation is not always met. Many varieties of English, for instance, allow (or even prefer) accusative case in conjoined structures in environments where nominative would be expected.


It is also not difficult to find apparent violations of the Law of the Coordination of Likes, as in (12), where each conjunct would be acceptable on its own, but they do not match under standard assumptions about syntactic categories.


Exceptions to the Coordinate Structure Constraint, especially those involving extraction only out of the second conjunct, are also well attested.


This tension between the apparently pervasive symmetry of coordination on the one hand and the occasional clear cases of asymmetry on the other hand is one of the reasons why coordination has been such a rich domain of study and interesting analytical puzzle for many decades. This article will present in more detail the facts that lead to this tension, as well as some of the major analyses that have been proposed.

1. Basic Properties of Coordination

How one understands the apparently conflicting evidence for symmetry and asymmetry in coordination plays a big role in determining the type of structure that one chooses to adopt, which, in turn, is a major factor in determining what types of structures are allowed in natural language syntax more generally. If one assimilates coordination to the more widely studied cases of subordination, in which structures are clearly right- (or left-) branching, then one conjunct c‑commands another and one would expect them to behave asymmetrically. Under such a view, the challenge then becomes how to account for the aspects of their behavior that appear symmetric. However, if one assumes that coordination is not right- (or left-) branching, and thus very different from subordination, the symmetric behavior is easier to account for, and it is the asymmetries that require some explanation. Given these relatively high stakes, we now look more closely at some of the better-studied cases where the tension between symmetry and asymmetry in coordination seems particularly clear.

1.1. Case and Agreement

In languages were case is marked overtly, conjoined nominal expressions generally carry the case that would be expected for the larger conjoined structure, and this case is realized symmetrically on each of the conjuncts. In (13), for instance, the pronouns he and I are conjoined, and they are each nominative, as would be expected given that [he and I] occupies the subject position.


Similar examples may be found in languages where all nominals are case-marked. Examples (14) and (15), for instance, show sentences from Tamil and Japanese, respectively, where the conjoined structure is a direct object and where each conjunct is marked for accusative.



The symmetric pattern in (13) through (15) appears to represent the general case; each conjunct receives the case that would be expected if the conjunct were to appear in that position on its own. Another type of symmetry sometimes occurs, though, in which each conjunct has the same case, but one that is not expected given the position of the conjoined phrase. This is seen in (16) for Bergen Norwegian, where each conjunct has accusative case, despite their being in subject position.


Alongside these types of symmetry, however, we find clear cases where case is allowed, or even required, to be distributed among the conjuncts asymmetrically. Tamil and Japanese, for instance, both allow the variants of (14)–(15) shown in (17) – (18), where only the final conjunct carries overt case marking.



The reverse is sometimes observed as well, where the second conjunct is nominative when accusative would be expected, as in English (19).


Similar phenom ena occur with subjects in some languages, where one of the conjuncts may have an “unexpected” case, as in Bergen Norwegian (20).


Examples of asymmetry may also be found in agreement systems, in that a head may agree with one or the other of the conjuncts. In Palestinian Arabic VSO clauses, for instance, the verb agrees in gender with the first conjunct, not the second, as seen in (21).


In Hopi, the asymmetry goes in the other direction, with agreement being sensitive to the second conjunct, not the first:


Cases of asymmetric agreement such as these may be viewed against the backdrop of symmetric agreement, in which the head agrees with features that the conjuncts have in common, as in (23), or with features associated with the larger conjoined phrase, as in (24a) and Spanish (24b).



These types of symmetric agreement appear to be more common cross-linguistically than asymmetric cases such as those in (21)–(22).

1.2. Coordination of Likes

Another area of tension between symmetry and asymmetry in coordination involves the categorial status of the conjuncts. It has been known for many decades that ill-formedness often results when each conjunct is of a different category, even though each would be possible within the sentence on its own. This was seen in (8) above, and additional examples are given here in (25) (from Peterson, 2004).


Notice that conjuncts may otherwise be very different from each other (e.g., in their internal structure), as long as they are of the same category:


This apparent requirement of categorial symmetry between the two conjuncts is known as the Law of the Coordination of Likes (LCL) (Williams, 1978).

Along with the basic facts motivating the LCL, such as those in (23), there are many apparent counterexamples, such as those in (27).


Although the conjuncts in each of the examples in (27) are clearly of different syntactic categories under a traditional analysis, they nonetheless appear to share a semantic type or role. Descriptively, at least, the LCL thus seems to allow either syntactic category or semantic role to be what counts for determining whether the conjuncts are alike.

1.3 Non-Constituent Coordination

Coordination is one of the traditional diagnostic tests for constituent structure. If a string x is able to conjoin with a similar string, this is taken to mean that x is a constituent, while if this is not possible, it suggests that x is not a constituent. This test is based on the fact that, in general, only conjuncts that are independently known to be constituents are allowed. Given the standard constituent structure in (28), for example, the coordinate structures shown in (29) are possible and those in (30) are not.




The distinctions between the acceptable sentences in (29) and the unacceptable ones in (30) are sharp, and the conclusions that they suggest about the structure of (28) are in line with evidence from many other diagnostic tests.

Coordination is not always such a reliable indicator of constituency, however. It is well known that the coordination test yields many false positive results, as in these examples within the VP:


The brackets in (31) show the conjuncts, but these conjuncts are not constituents under traditional analyses of clause structure. Coordination thus presents an interesting puzzle: it appears to obey constituent structure very strictly in cases such as (29)–(30), but is able to flout it easily in cases such as (31).

2. Coordination and Extraction

2.1. The Coordinate Structure Constraint and Across-the-Board Extraction

The topic of extraction out of coordinate structures has been a particularly fertile one for research, keeping scholars occupied for decades and providing interesting evidence about the nature of both extraction and coordination. The most basic fact in this domain is that conjuncts may not be extracted, as seen in (32).


Extraction out of conjuncts, however, is permitted, though in the general case, this is limited to cases where extraction is out of all conjuncts, as seen in (33).


Facts such as those in (32)–(33) were originally discussed by Ross (1967) and motivated his Coordinate Structure Constraint. Since that time, the phenomenon of simultaneous extraction out of all conjuncts, as in (33c), has been known as Across-the-Board (or ATB) extraction.

The prohibition on conjuncts themselves being extracted, as in (32), appears to be exceptionless, but the ATB restriction, as in (33), shows more nuanced behavior. This restriction is at its strongest when the conjuncts are semantically independent, in the sense that their relative order does not matter. This is the case in (33). When the order of the conjuncts does matter, however, such as when there is an implicit temporal or causal relation between them, non-ATB extraction is possible, as in (34).


ATB extraction also shows interesting behavior even in those cases where the effect is very robust, as in (33). By definition, ATB extraction involves having a gap in every conjunct, but the position of the gap seems to affect acceptability. Strict parallelism (e.g., all subject gaps or all object gaps) yields the most acceptable cases, as in (33c) or (35).


This is especially true when conjuncts begin with a subject gap, as in (35). When such a conjunct is conjoined with a conjunct with a gap in a different position, noticeable degradation occurs, as seen in (36).


The effect is strongest when the subject gap occurs directly after the conjunction, as in (36b). For conjuncts that do not begin with a subject gap, there is much less of a preference for parallelism, so having conjuncts with gaps in different positions is still relatively acceptable.


2.2. Coordinated wh-phrases

Another interesting fact about coordination and extraction is that it is possible for the extracted wh-phrases themselves to be coordinated, as in (38).


Such wh-questions may seem unremarkable, but they are often straightforward counterexamples of the Law of the Coordination of Likes. Consider the contrast in (39), for example.


Though wh-phrases appear to allow for a loosening of the LCL, they still obey the subcategorization requirements of the head verb. The verb sing, for instance, is able to appear with or without a direct object, as in (40), and it is this property that permits it to take coordinated wh-phrases as in (39a).


Verbs that do not have this property do not allow coordinated wh-phrases of this type, as seen in the following example


In English, coordinated wh-phrases typically involve conjoined arguments of the same type (both receiving the same theta-role), as in (38a), conjoined adjuncts, as in (38b), or an argument and an adjunct, as in (39a). Some other languages extend this further, by allowing for conjoined arguments of different types, where both are obligatory and each receives a different theta-role. This may be seen in the Russian example in (42).


This is not possible with coordination of non-wh-phrases, as seen in (43).


3. Other Topics Within Coordination

3.1. Gapping

The term Gapping refers to structures in which a clause is conjoined with something that is like a clause, but in which material in the middle is missing. A characteristic example is given in (44), with the missing material represented by “__.”


The second conjunct here is not a full clause, but it is understood as if it were (i.e., John ate potatoes). The verb in this case is “gapped,” leaving the two remnants John and potatoes.

The gapped string in Gapping sentences typically involves at least the verb, and it does not need to be a constituent, as seen in (45).


In all four of these cases, the gapped string (should have been eating, will donate money, wants to try cooking, and Mary eats) is not a constituent, under standard assumptions. These examples also demonstrate that the gapped string can be relatively long and complex. Despite the diverse nature of the gapped strings that we have seen, they do all contain some verbal material. This appears to be a requirement:


In (46), the gapped string (Mary) contains no verb.

Although the gapped string does not need to be a constituent, the remnants typically are. Put more precisely, there appears to be a strong preference for exactly two remnants, with one on each side of the gapped string and each one a constituent. Unacceptable cases where this condition is not obeyed are given in (47).


The constraints on Gapping mentioned so far are not exhaustive, however; many additional requirements have been discussed in the literature. Embedded clauses, for instance, are well-formed remnants when they are complements, as in (48a), but not when they are subjects, as in (48b).


The source of unacceptability in (48b) appears to be the clausal subject, since everything else about the sentence, including the gapped string (would surprise), seems to be well formed.

It has also been noted that Gapping resists cases where either the gapped string or its antecedent are in an embedded clause. This is shown in (49a), where the gap is in the embedded clause, and in (49b), where the antecedent is in the embedded clause.


Crucially, (49b) here must be understood such that John __ potatoes is conjoined with I think that Mary eats beans (and not just with Mary eats beans, in which case it is perfectly acceptable). Example (49a) is independently ruled out by the requirement that the remnants be constituents (I think that John is not), but (49b) obeys this requirement, thus suggesting that the embedding restriction is real and independently motivated.

Gapping appears to be a very widespread phenomenon cross-linguistically, but an interesting correlation has been observed between basic word order and type of Gapping. Verb-medial languages, such as English, and verb-initial languages generally have forward Gapping, where it is the second conjunct that has missing material, as in (50a) and (50b), whereas verb-final languages typically have backward Gapping, where the first conjunct is the one with missing material, as in (50c).


3.2. Right Node Raising

The term Right Node Raising (RNR) refers to structures in which two sentences that share a final constituent are conjoined, and this final constituent appears in a right-peripheral position.


That the final constituent must in fact be a constituent is suggested by examples such as (52).


The examples in (53) show that this constituent must be the rightmost element in the conjoined clauses.


Right Node Raising also obeys the Across-the-Board property. That is, right-peripheral constituent must be missing from all of the conjuncts:


Right Node Raising has prosodic preferences as well. Sentences in which the right-peripheral constituent is not sufficiently heavy generally sound degraded:


Although RNR is often discussed in relation to coordination, it is not limited to coordinate structures:


3.3 Comitative Coordination

In some languages, it is possible for a verb that has a morphologically plural argument and a comitative phrase (e.g., a prepositional phrase headed by with) to be interpreted so that the plural-marked argument is understood as a singular argument coordinated with the comitative argument. In Hausa (57a) and Spanish (57b), for example, the subject is marked as first-person plural, but it is understood as I coordinated with the comitative argument.


Similarly in Dakota (58), the object is marked as first-person dual, and it is understood as you and me.


This phenomenon is known as comitative coordination. It occurs in a wide and typologically diverse range of languages, but clearly not in English, as may be seen in the interpretation given to sentences like (59).


Crucially, there must be at least three people in (59) that went to the movies: John and the two or more people that he accompanied. If English were to allow the comitative coordination meaning, then (59) could be interpreted as John and I went to the movies.

4. Critical Analysis of Scholarship

4.1. The Source of Coordination: Derived or Phrasal

One of the earliest ideas in generative research on coordination is that sentences like (60) are derived from sentences like (61) through a process of ellipsis.



Assuming that ellipsis processes are motivated independently in the grammar and that they allow for repeated material to not be pronounced, then deriving (60) from (61) appears very natural. Such a process might seem less natural in cases like (62), which would have to be derived from (63) by not pronouncing the first instance of the material.



In addition, the idea that sentences like (64) would have to be derived from the semantically deviant (65) has led to widespread skepticism that all coordination is uniformly the result of ellipsis.



Nonetheless, ellipsis analyses of certain types of coordination, especially those that have a marked prosodic pattern, have remained popular. Johnson (2007), te Velde (2006), and Wilder (2008) are examples of this.

Another early idea in the analysis of coordination is that coordinate structures are base-generated as such, with sentences such as (60), for example, relying on a rule such as (66).


This type of analysis eliminates concerns about the directionality of ellipsis (cf. (62)–(63)) and can generate a sentence like (64) directly, without deriving it from (65). One traditional objection to this type of analysis, though, is that if one assumes that subjects of active clauses are base-generated and that subjects of passives are derived, then it is difficult to see how an active VP and a passive VP, both sharing a common subject, could be coordinated. Such sentences are clearly possible, however, as seen in (67).


Clearly, whether sentences like these present a problem for a phrasal analysis of coordination depends on one’s analysis of passive clauses and of clause structure. In analyses in which passive subjects are not derived, (67) is unlikely to pose a problem (see, for example, Gazdar, 1981; and Sag, Gazdar, Wasow, & Weisler, 1985). Even in analyses in which passive subjects are derived, the particular clause structure that one adopts may allow for a solution. Burton and Grimshaw (1992) and McNally (1992), for instance, show that, by adopting the notion of VP-internal subjects, where all subjects are derived, the problem posed by (67) essentially disappears. In effect, then, current analyses either treat all subjects as base-generated or treat all subjects as derived, so sentences like (67) are no longer taken to be problematic.

The historic tension between derived and phrasal analyses of coordination does not occupy a central position in current studies, as there is a widespread assumption that aspects of both approaches (specifically ellipsis and conjunction of phrases) are required, but the tension nonetheless still arises occasionally, particularly with regard to analyses of specific phenomena, such as Gapping and Right Node Raising.

4.2. The Structure of Coordination: Headed, Non-Headed, Parallel

Scholarship on the basic structure of coordination may be divided into three types. First, there are proposals in which the coordinate structure is headed by the conjunction, and the conjuncts fit into this structure in a way that conforms with X’-theory. The appeal of such an analysis is understandable: if the coordinate structure is a phrase, as seems undeniable, then we would want it to obey the properties of phrases that we observe more generally. The appeal is heightened by the fact that the conjunction does in fact display head-like properties (e.g., it must be a single word), and the fact that the conjuncts are themselves typically phrases, as one would expect if they are in a specifier or complement relation to the head. Camacho (2003), Munn (1993), Johannessen (1998), and Zhang (2010) present analyses along these general lines, while Borsley (2005) offers a critique.

Second, some analyses have proposed that the coordinate structure is not headed and that the conjuncts instead share their features with their mother node. This type of analysis also has a clear, intuitive appeal; it has long been noted that a coordinate structure is of the same syntactic category as its conjuncts and in these proposals, that fact plays a central role in the formal analysis. Gazdar (1981), Sag et al. (1985), and Chaves (2012) are some of the most prominent instantiations of this type of analysis.

Third, there are proposals that claim that coordinate structures cannot be expressed using traditional phrase structure and that three-dimensional or multidominant structures are required instead. In such analyses, the conjuncts exist in parallel within the larger sentence structure, with neither one dominating or preceding the other. These analyses capture the intuition that each conjunct would be able to stand on its own within the larger structure. Well-known proposals along these lines include Williams (1978), Goodall (1987), and Citko (2005, 2011).

Further Reading

  • Aoun, J., Benmamoun, E., & Sportiche, D. (1994). Agreement, word order, and conjunction in some varieties of Arabic. Linguistic Inquiry, 25, 195–220.
  • Chaves, R. P. (2012). On the grammar of extraction and coordination. Natural Language and Linguistic Theory, 30, 465–512.
  • Citko, B. (2005). On the nature of merge: External merge, internal merge, and parallel merge. Linguistic Inquiry, 36, 475–496.
  • Citko, B., & Gračanin-Yuksek, M. (2013). Towards a new typology of coordinated wh‑questions. Journal of Linguistics, 49(01), 1–32.
  • Citko, B. (2011). Symmetry in syntax: Merge, move, and labels. Cambridge, U.K.: Cambridge University Press.
  • Dalrymple, M., & Kaplan, R. M. (2000). Feature indeterminacy and feature resolution. Language, 76, 759–798.
  • Goodall, G. (1987). Parallel structures in syntax: Coordination, causatives, and restructuring. Cambridge, U.K.: Cambridge University Press.
  • Hankamer, J. (1979). Deletion in coordinate structures. New York: Garland.
  • Haspelmath, M. (Ed.). (2004). Coordinating constructions. Amsterdam, Netherlands: John Benjamins.
  • Johannessen, J. B. (1998). Coordination. Oxford: Oxford University Press.
  • Johnson, K. (2007). Gapping. In M. Everaert & H. van Riemsdijk (Eds.), The Blackwell companion to syntax (pp. 407–435). Malden, MA: Blackwell Publishing.
  • Peterson, P. G. (2004). “Coordination: Consequences of a lexical-functional account.” Natural Language and Linguistic Theory, 22, 643–679.
  • Phillips, C. (2003). Linear order and constituency. Linguistic Inquiry, 34, 37–90.
  • Ross, J. R. 1967. Constraints on variables in syntax. Doctoral diss., MIT, Cambridge, MA.
  • Sabbagh, J. (2014). Right node raising. Language and Linguistics Compass, 8(1), 24–35.
  • Sag, I., Gerald G., Thomas W., & Weisler, S. (1985). Coordination and how to distinguish categories. Natural Language and Linguistic Theory, 3, 117–171.


  • Aoun, J., Benmamoun, E., & Sportiche, D. (1994). Agreement, word order, and conjunction in some varieties of Arabic. Linguistic Inquiry, 25, 195–220.
  • Arden, A. H. (1954). A progressive grammar of common Tamil. Madras, India: Christian Literature Society.
  • Borsley, Robert D. (2005). Against conjP. Lingua, 4, 461–482.
  • Burton, S., & Grimshaw J. (1992). Coordination and vp-internal subjects. Linguistic Inquiry, 2, 305–313.
  • Camacho, J. (2000). “Structural Restrictions on Comitative Coordination.” Linguistic Inquiry, 31, 366–375.
  • Camacho, J. (2003). The structure of coordination: Conjunction and agreement phenomena in Spanish and other languages. Dordrecht, Netherlands: Kluwer Academic Publishers.
  • Chaves, R. P. (2012). On the grammar of extraction and coordination. Natural language and linguistic theory, 30, 465–512.
  • Citko, B., & Gračanin-Yuksek, M. (2013). Towards a new typology of coordinated wh‑questions. Journal of Linguistics, 49(01), 1–32.
  • Gazdar, G. (1981). Unbounded dependencies and coordinate structure. Linguistic inquiry, 12, 155–184.
  • Goodall, G. (1987). Parallel structures in syntax: Coordination, causatives, and restructuring. Cambridge, U.K.: Cambridge University Press.
  • Johannessen, J. B. (1998). Coordination. Oxford: Oxford University Press.
  • Johnson, K. (2007). Gapping. In M. Everaert & H. van Riemsdijk (Eds.), The Blackwell companion to syntax (pp. 407–435). Malden, MA: Blackwell.
  • Kazenin, K. (2002). On coordination of wh-phrases in Russian. Ms., Tubingen University, Germany, and Moscow State University, Russia.
  • Larsen, A. B., & G. Stoltz. (1912). Bergens bymaal. Kristania (Oslo): Bymålslaget/H. Aschehoug.
  • Martin, S. E. (1975). A reference grammar of Japanese. New Haven, CT: Yale University Press.
  • McNally, L. (1992). VP coordination and the VP-internal subject hypothesis. Linguistic Inquiry, 23, 336–341.
  • Munn, A. (1993). Topics in the syntax and semantics of coordinate structures. Doctoral diss., University of Maryland, College Park.
  • Peterson, P. G. (2004). Coordination: Consequences of a lexical-functional account. Natural Language and Linguistic Theory, 22, 643–679.
  • Ross, J. R. (1967). Constraints on variables in syntax. Doctoral diss., MIT, Cambridge, MA.
  • Sag, I., Gazdar, G., Wasow, T., & Weisler, S. (1985). Coordination and how to distinguish categories. Natural Language and Linguistic Theory, 3, 117–171.
  • Schwartz, L. (1988). Conditions on verb-coded coordinations. In M. Hammond, E. Moravcsik, & J. Wirth (Eds.), Studies in syntactic typology (pp. 53–73). Amsterdam, Netherlands: John Benjamins.
  • te Velde, J. (2006). Deriving coordinate symmetries: A phase-based approach integrating select, merge, copy, and match. Amsterdam, Netherlands: John Benjamins.
  • van Oirsouw, R. (1987). The syntax of coordination. London: Croom Helm.
  • Wilder, C. (2008). Shared constituents and linearization. In Johnson, K. (Ed.), Topics in ellipsis (pp. 229–258). Cambridge, U.K.: Cambridge University Press.
  • Williams, E. (1978). Across the board rule application. Linguistic Inquiry, 9, 31–43.
  • Zhang, N. N. (2010). Coordination in syntax. Cambridge, U.K.: Cambridge University Press.