Head-directionality parameter

From Wikipedia, the free encyclopedia
  (Redirected from Head directionality parameter)
Jump to: navigation, search

In linguistics, the head-directionality is a proposed parameter that classifies languages according to whether they are head-initial (the head precedes its complements) or head-final (the head follows its complements). A head is an element that determines the category of a phrase. For example, if a verb is the head a phrase, then the phrase is a verb phrase. While some languages are consistently head-initial or head-final at all phrasal levels (such as Japanese), other languages use a mixture of the two parameters (such as German).

The head-directionality parameter can be examined at the phrasal level to determine whether the phrase is head-initial or head-final. Following this scheme, four languages (English, German, Japanese, and Gbe) are examined on their surface syntax to explore the head-directionality parameter within each phrase. Since many languages are not always consistent in their head-directionality parameter across different phrases in a single language, many theories have been proposed to explain the inconsistencies. The surface true and statistical theories examine the surface syntax of a given language to determine the head-directionality parameter. The antisymmetry approach examines the deep syntax and proposes that every language is head-initial in their underlying structures.


The idea that syntactic structures reduce to binary relations was first introduced by Lucien Tesnière within the framework of dependency theory. This idea was taken up in Noam Chomsky's influential principles and parameters theory, which was developed in the 1980s.

Dependency theory: Tesnière 1959[edit]

Dependency theory, developed in the 1960s, posits two structures that differ relative to the placement of the structurally governing element (head).[1]

  • Centripetal structures: heads precede their dependents.
  • Centrifugal structures: heads follow their dependents.

Dependents include adjuncts, specifiers and complements.

Principles & parameters theory: Chomsky 1981[edit]

Noam Chomsky's Principles and Parameters theory in the 1980s [2] introduced the idea that a small number of innate principles are common to every human language (e.g. phrases are oriented around heads), and that these general principle are subject to parametric variation (e.g. the relative order of head can differ). This has come to be known as the "principles and parameters" theory. In this theory, the dependency relation between heads, complements, specifiers, and adjuncts is regulated by X-bar theory. The complement is sister to the head, and can be ordered in one of two ways. A head-complement order — a special case of Tesnière's centripetal structure — is called a head-initial structure. A complement-head order — a special case of Tesnière's centrifugal structure — is called a head-final structure. Within the principles & parameters theory, there is a head-directionality parameter that distinguishes languages according to whether they are head-final or head-initial, as outlined below:

  • Head-initial phrases: heads precede their complements.
  • Head-final phrases: heads follow their complements.

The head-directionality parameter depends on X-bar theory that Jackendoff [3] proposed in the 1970s, which differentiates the theory with regard to Tesnière's theory. X-Bar theory requires phrases to be endocentric (based on a head), with a complement and specifier.[4] Under this theory, a phrase would have some internal structure governing the relationships between the head, complement, and specifier, and thus have head-directionality relative to the two arguments of the complement and the specifier.

Head-directionality is used to classify languages according to their linguistic typology and to examine relationships within and across languages. Some linguists use statistical analysis to construct a model for language typology. However, directionality is not a perfect descriptor as the data below will show. Very few, if any, languages are entirely one direction or the other. Thus linguists have come up with a number of theories to explain the inconsistencies. Some posit that the word order of a language represents the underlying relations in the language or, like Kayne, use phrasal movement to explain directionality.

Head directionality versus word order[edit]

Head directionality is related to issues in word order typology. One of the chief parameters in this, concerns the normal order of subject, verb, and object within a clause (SVO, VSO, etc.) The head-directionality parameter, however, considers positioning within a phrase – the spatial relationship between the head of a phrase, and dependents such as adjuncts, specifiers, and complements. This concerns all types of phrases, such as noun phrases, determiner phrases, prepositional/postpositional phrases and others listed below. All of these phrases have some sort of internal order. By looking at how the heads of different phrases interact with their dependents, one can theoretically determine if a language as a whole is head-final or head-initial.

Consider the example in (1), which shows that English can be analyzed as a language in which heads precede their complements.[5] In the complementizer phrase (CP) in (1a), the complementizer (C) precedes its tense phrase (TP) complement. In the tense phrase in (1b), the tense-marking element (T) precedes its verb phrase (VP) complement. In the verb phrase in (1c), the verb (V) precedes its two complements, namely the determiner phrase (DP) the book and the prepositional phrase (PP) on the table. In (1d), a picture is analyzed here as a determiner phrase, the determiner (D) a precedes its noun phrase (NP) complement, while in (1e), the preposition (P) on precedes its DP complement your desk.

(1) English: head-initial
      You will see 'that the girl will put a picture on your desk'.
      a. CP: [CP [C that ]  [TP the girl will put a picture on your desk ] ]
      b. TP: [TP [T will ]  [VP put a picture on your desk ] ]
      c. VP: [VP [V put ]   [DP a picture ] [PP on your desk ] ]
      d. DP: [DP [D a ]     [NP picture ] ]
      e. PP: [PP [P on ]    [DP your desk ] ]
      (Sportiche, Koopman & Stabler 2014 97)

Based on this type of analysis, English and Romance languages are often given as examples of head-initial languages, whereas Japanese and Basque are often given as examples of head-final languages. However, this simple dichotomy runs into major problems when used to classify entire languages by this parameter. Many languages are not exclusively head-initial or head-final within all areas of their grammar. Even languages that are considered strongly head-initial contain head-final phrases, and languages that are considered strongly head-final contain head-initial phrases. English, for example, is largely considered head-initial, but as will be shown below, it can have head-final adjective phrases. Some different theoretical approaches to these issues will be discussed later in this article.

Head position in specific types of phrase[edit]

This section will examine the surface syntax of specific types of phrases within languages that have been described as being either head-initial, head-final or both. It will focus primarily on the relation between the head of the phrase and its complements, rather than the specifier or adjuncts. Adjuncts are slightly more flexible than complements in terms of order and, like specifiers, have a different relation to the head of the phrase.

Lexical categories[edit]

The lexical categories considered here are nouns, verbs, adjectives, and prepositions/postpositions. Each of these can appear as the head of a corresponding type of phrase: noun phrase, verb phrase, adjective phrase, or adpositional phrase.

VP: relative order of the verb[edit]

See also: Verb phrase

A verb phrase contains a lexical head, namely the verb, and its complements. The verb takes complements such as determiner phrases and complementizer phrases. This section will examine the verb and its complements in order to explore whether the verb phrase is head-initial or head-final in the example languages.

Example (2) below shows the placement of the verb in a verb phrase in English, showing that V precedes its complements.[6] The complement that the verb precedes is the DP an apple.

(2) English VP structure
2) English
     a.  eat an apple
         (Fukui 1993 4: (3a))
     b.  [VP [V eat] [[DP an apple]]

Switching the order to "[VP[DP an apple] [[V eat]]" would be ungrammatical. This shows that for VPs, English is often head-initial: the head (V) occurs before its complements.

In German, a verb may occur after its complements. Example (3) illustrates a head-final structure since the verb finden follows its complement es.[7] The auxiliary verb werde is analyzed as a tense so it sits in the structure above the verb as indicated by the tree diagram on the right.

(3) German VP structure
3) German
     a. Ich werde es finden'
        I   will  it find
        'I will find it.'
        (Dopke 1998 6: (2))
     b. [TP [DP Ich] [T' [T werde]] [VP [DP es] [V finden]]]]

Japanese appears to be head-final for VPs because the complement occurs before the V head.[6] In (5), the complement ringo-o (a DP) proceeds the verb tabe. Ru is a tense marker (T) that attaches to the verb in accordance with the rules for tense phrases.

(4) Japanese VP structure
4) Japanese
     a. ringo-o   tabe-ru
        apple-acc eat-nonpast
        'eat an apple'
        (Fukui 1993 4: (3b))
     b. [TP [VP [DP ringo-o] [[V tabe]] [[T ru]]

Examples (5) and (6) show that in Gbe, the verb may appear after or before its complement.[8] The verb for "use" appears after its DP complement in (5) and before its DP complement in (6). Thus in Gbe, there can be both head-initial or head-final structures for the verb phrase in their surface syntax.

(5) Gbe (head-final) VP structure
5) Gbe
     a. Kɔ̀jó tó     àmí lɔ́  zân
        Kojo Imperf oil Det use-NR
        'Kojo is using the oil.'
        (Aboh 2001 1: (2a))
     b. [TP [DP Kɔ̀jó] [T' [T tó] [VP [DP àmí lɔ́][V zân]]]
(6) Gbe (head-initial) VP structure
6) Gbe
     a. Kɔ̀jó nɔ̀  zán      àmí lɔ́
        Kojo hab use-perf oil det
        'Kojo habitually used the oil/Kojo habitually uses the oil.'
        (Aboh 2001 1: (2c))
     b. [TP [DP Kɔ̀jó] [T'[T nɔ̀] [VP [V zán] [DP àmí lɔ́]]]]

It has been debated whether example (5) is due to object movement to the left side of the verb [9] or whether the lexical entry of the verb simply allows head-initial and head-final structures.[10] This debate considers the underlying structure of the sentences and posits movement. However, since only the surface syntax is considered in this typology, it is evident that both head-initial and head-final structures are present in this language.

NP: relative order of the noun[edit]

A noun phrase has a noun as its head. The noun can take different complements such as CPs and PPs shown in the data below. The noun can be ordered in different ways with its complements so that head-initial or head-final structures are both theoretically possible.

In example (7), the noun girl takes a complementizer phrase as a complement. This CP occurs after the head N. This is evidence that the noun phrase in English is head-initial.[11] The tree structure on the right also illustrates this.

(7) English NP structure
7) English
     a. He married a girl who is from Texas.
        (Smith 1964 6: (14))
     b. [NP [N girl] [CP who is from Texas]]]

In (8), the CP der den Befehl überbrachte follows the head N boten, demonstrating that within NPs, German is head-initial as well.[12] This contrasts with the head-final pattern of German VPs showing that within a language there may be different head parameters for different phrases.

(8) German NP structure
8) German
     a. Man beschimpfte den Boten,    der den Befehl  überbrachte
        one insulted    the messenger who the command delivered
        'The messenger was insulted who delivered the command.'
        (Berthold 2005 6: (11))
     b. [NP [N Boten] [CP der den Befehl überbrachte]]]

In example (9), the noun koogi is preceded by New York-de-no, a prepositional phrase that is a complement of N. This shows that Japanese can be head-final in noun phrases.[13]

(9) Japanese NP structure
9) Japanese
     a. John-no  kinoo-no      New York-de-no  koogi
        John-gen yesterday-gen New York-in-gen lecture
        'John's lecture in New York yesterday'
        (Fukui 1993 16: (27a))
     b. [NP [PP New York-de-no} [N koogi]]]

In example (10), from Gbe, the noun "sketch" (ɖìdè) occurs after its complement.[14] This suggests that the noun phrase in Gbe is head-final. The noun phrase in Gbe contains a different head order constraint than the verb phrase since the verb phrase can be both head-final or head-initial.

(10) Gbe NP structure
10) Gbe
     a. Kɔ̀kú sín  ɖìdè   lɛ̀
        Koku case sketch NBR
        'sketches of Koku'
        (Brousseau & Lumsden 1992 3: (4))
     b. [NP [KP Kɔ̀kú sín] [[N ɖìdè]]

AP: relative order of the adjective[edit]

Adjective phrases contain an adjective head, and can take as a complement, for example, an adverb phrase (AdvP) or prepositional phrase.

In example (11) below, the adjective happy does not have a complement. The entire adjective phrase acts acts as an adjunct to NP. However, it is generally accepted that an adjective phrase that serves as an adjunct that occurs before a noun has a head-final structure.[15]

(11) English (head-final) AP structure
11) English
     a. extremely happy person
         (Sadler & Arnold 28: (58a))
     b. [NP [AP [Adv extremely] [[A happy]] [[NP person]]

Example (12), however, shows that modifiers such as the prepositional phrase about her work follow the head of the adjective phrase (happy), suggesting that there is some fluidity as to where different kinds of modifiers appear in adjective phrases.[15] Since the complement PP occurs after the adjective in (12), this type of adjective phrase that occurs after the noun, contains a head-initial structure.

(12) English (head-initial) AP structure
12) English
     a. person happy about her work
         (Sadler & Arnold 28: (58b))
     b. [NP [N person] [AP [A happy] [PP about her work]]]

In (13), the German adjective stolze occurs after its complement PP seine Kinder.[16] This would suggest that the adjective phrase in German is head-final.

(13) German (head-final) AP Structure
13) German
     a. der auf seine Kinder stolze vater
        the of his children proud+Agr father
        'The father (who is) proud of his children'
        (Hinterhölzl 2010 4: (8a))
     b. [AP [PP auf seine Kinder] [[A stolze]]

However, (14) shows the opposite pattern: the adjective stolz occurs before its complement, demonstrating a head-initial ordering.[17] This indicates that German allows for head-final or head-initial structures in adjective phrases.

(14) German (head-initial) AP Structure
14) German
     a. weil er stolz auf seine Kinder ist
        since he proud of his children is
        'Since he is proud of his children'
        (Hinterhölzl 2010 5: (8d))
     b. [AP [A stolz] [[PP auf seine Kinder]]

In Japanese, in example (15), the intensifier (Int) ni-juu-meetoro ("twenty meters") precedes the A head takai ("tall"), indicating that within an adjective phrase, Japanese uses the head-final structure, similar to the other phrasal types in Japanese covered thus-far.[18]

(15) Japanese AP Structure
15) Japanese
     a. Kono biru-wa      ni-juu-meetoru takai
        this building-TOP two-ten-meter  tall
        ‘This building is twenty meters taller.’
        (Watanabe 2011 11: (29))
     b. [AP [Int ni-juu-meetoru] [[A takai]]

In the following example, Gbe follows a head-initial pattern because the head precedes the intensifier tàùú.[19]

(16) Gbe AP Structure
16) Gbe
     a. ún   mɔ̀n         àǔn  yù      tàùú àtɔ̀n éhè lɔ̀       lɛ́
        Isg see-Perf    dog black    Int Nral Dem Spf[+def] Num
        'I saw these specific three really black dogs'
        (Aboh 2004 100: (44a))
     b. [AP [A ] [[Int tàùú]]

PP: relative order of the pre/post-position[edit]

Adpositional phrases (PPs) are called prepositional phrases if they are head-initial (i.e. headed by a preposition), or postpositional phrases if they are head-final (i.e. headed by a postposition). The complement, in this analytical scheme, is a determiner phrase.

As example (17) shows, English has prepositions rather than postpositions, that is, its adpositional phrases are head-initial. Here the P head of precedes the complement DP eligible voters.[20] This continues the general pattern of head-initial phrases in English.

(17) English PP structure
17) English
      a. A majority of eligible voters prefer not to vote
         (Gillon 1992 15: (25))
      b. [PP [P of] [DP eligible voters]]]

German, however, allows adpositional phrases with both head-final and head-initial structures. In (18), the postposition hinauf occurs after its complement das Dach.[21]

(18) German head-final PP structure
18) German
     a. Die Schnecke kroch das Dach hinauf
        the snail crept the.ACC roof up
        The snail crept upward along the roof
        (van Riemsdijk 2007 1: (1a))
     b. [PP [DP das Dach] [[P hinauf]]

In (19), the preposition auf occurs before its complement den Tisch. This shows that the German PP can also be head initial.[22] This continues the trend where German allows different head-parameters within phrases.

(19) German head-initial PP structure
19) German
     a. Peter legt das Buch auf den Tisch
        Peter puts the book on the.ACC table
        Peter puts the book on the table
        (van Riemsdijk 2007 3: (4b))
     b. [PP [P auf] [[DP den Tisch]]

As with previously considered phrase types, Japanese uses a head-final structure in its PPs. In example (20), the P head ni follows the DP Takasu-mura.[23]

(20) Japanese PP structure
20) Japanese
      a. Boku-ga Takasu-mura-ni    sunde-iru
         I-nom   Takasu-Village-in live-pres
         'I live in Takasu Village.'
         (Vinka 2009 4: (11b))
      b. [PP [DP Takasu-mura] [[P -ni]]

In example (21), from Gbe, the P xlán proceeds the DP Àsíbá, giving evidence that Gbe is head-initial in PPs.

(21) Gbe PP Structure
21) Gbe
      a. Kòfi     zé        kwɛ́      xlán  Àsíbá"
         Kofi   take-Pref    money    prep  Asiba
         'Kofi sent money to Asiba'
         (Aboh 2004 117: (4a))
      b. [VP [DP Kòfi] [V' [V zé] [DP kwɛ́] [PP [P xlán] [DP Àsíbá]]]]

Functional categories[edit]

Functional categories in this typology are determiners, complementizers, tenses, and aspects. These functional categories are also known to be closed class because words are not easily added. Words in this class typically carry grammatical meaning.

DP: relative order of the determiner[edit]

The determiner phrase is a proposed phrase that contains a determiner as the head of the phrase rather than a noun. This phrase was proposed under generative syntax[24] and not all theories of syntax agree that this phrase exists.[25] There continues to be a debate over the existence of this phrase. The determiner phrase can still be examined to determine whether or not it is head-final or head-initial.

In English (22), the head any proceeds the NP book, suggesting that English is head-initial within the determiner phrase.[26] The tree of the right also depicts this relationship since the head is on the left compared to the complement.

(22) English DP structure
22) English
      a. any book
         (Takahashi 2002 2: (2c))
      b. [DP [D any] [[NP book]]

In (23), the D head der ('the'), precedes the NP Mann in der mann giving evidence that German is head-initial within DPs.[27] The tree structure between German and English is very similar as both languages contain head-initial determiner phrases.

(23) German DP structure
23) German
      a. der Mann, der in seinem Büro  arbeitet
         the man   who in his   office works
         'the man who works in his office'
         (Bianchi 1999 251: (111))
      b. [DP [D der] [NP [N Mann] [CP der in seinem Büro arbeitet]]]

In (24) below, Japanese is once again shown with a head-final structure, with the D head mo ('any') following the NP dare ('person').[26] The trees for Japanese and English are exactly the opposite since they have the exact opposite head parameter.

(24) Japanese DP structure
24) Japanese
      a. dare mo
         person any
         (Takahashi 2002 2: (3b))
      b. [DP [NP dare] [[D mo]]

In example (25), Gbe is shown with head-final structure, as the D head ɖé follows the NP complement àvɔ̀ àmàmú màtàn-màtàn.

(25) Gbe DP Structure
25) Gbe
   a. Asíbá   xɔ̀        àvɔ̀   àmàmú  màtàn-màtàn        ɖé
      Asiba buy-Perf   cloth   green    odd             Spf[-def]
      'Asiba bought a specific ugly green cloth'
      (Aboh 2004 100: (44b))
   b. [DP [D ɖé] [[NP àvɔ̀ àmàmú màtàn-màtàn]]

CP: relative order of the complementizer[edit]

The complementizer phrase contains either an covert C head, or an overt C like "that" in English, and its complements are TPs. This section will examine the head-directionality parameter within the complementizer phrase.

Example (26) shows how in English the complementizer that proceeds the TP Mary did not swim, giving further evidence for the head-initial property in English .[28]

(26) English CP structure
26) English
      a. That Mary did not swim
         That Mary did not swim
         'That Mary did not swim.'
         (Sells 1995 4: (3)))
      b. [CP [C That] [TP [DP Mary] [TP did] [VP [Neg not] [[VP swim]]

Like English, in German, the complementizer Daß proceeds the TP Lisa a flower planted has in example (27), and is thus head-initial.[29]

(27) German CP Structure
27) German
      a. daß Lisa eine Blume gepflanzt hat
         that Lisa a flower planted has
         'that Lisa planted a flower'
         (Kathol 2001 1: (2))
      b. [CP [C daß] [[TP Lisa eine Blume gepflanzt hat]]

In example (28), the complementizer to follows the TP Mary-non swim-neg-past ('Mary did not swim'), suggesting once again, that Japanese is head-final.[28]

(28) Japanese CP structure
28) Japanese
      a. Mary-ga  oyog-ana-katta-to
         Mary-non swim-neg-past-comp
         'That Mary did not swim'
         (Sells 1995 4: (4))
      b. [CP [TP [VP [VP [DP Mary-ga] [[V oyog]] [[Neg ana]] [[T katta]] [[C to]]

In the example below, Gbe follows a head-initial pattern in relation to complementizers. The C ɖé proceeds the tense phrase Dòsà gbá xwé ɔ̀ ɔ̀.

(29) Gbe CP Structure
29) Gbe
   a. ɖé   Dòsà    gbá      xwé   ɔ̀        ɔ̀
      that Dosa build-Perf house Spf[+def] DetCL
      'That Dosa built the house'
      (Aboh 2004 348 v:f)
   b. [CP [C ɖé] [[TP Dòsà gb xwé ɔ̀ ɔ̀]] 

TP: relative order of the tense[edit]

Tense is the head of a tense phrase. The tense expresses time such as the present, past, or future. This section will examine where the T exists in relation to its complement VP.

In (30), T precedes its complement VP: the tense did occurs before the verb swim. The tree diagram (30) shows that English has a head-initial parameter for the tense phrase, since T sits to the left of the VP.[28]

(30) English TP structure
30) English
      a. That Mary did not swim
         That Mary did not swim
         'That Mary did not swim.'
         (Sells 1995 4: (3)))
      b. [TP [DP Mary] [T' [T did] [VP swim]]]

The German tense precedes its complements in (31).[30] The tense werden precedes its VP complement jeden Brief beantworten den wir bekamen. Thus, German has head-initial structures for the complementizer and tense phrase, just as is seen above in English.

(31)German TP Structure
31) German
     a. Wir werden jeden Brief beantworten den wir bekamen
        We   will  every letter  answer   that we received 
        We will answer every letter  that we received. 
        (Kratzer 1998 15: (41b))
     b. [TP [T werdern] [[VP jeden Brief beantworten den wir bekamen]]

In example (32), Japanese's TP also has a head-final structure. The past tense katta occurs after the verb. In tree diagram (32), the T occurs to the left of the VP. Since the VP is a complement of T and the VP occurs on the left, Japanese has a head-final structure in its TP.[28]

(32) Japanese TP structure
32) Japanese
      a. Mary-ga  oyog-ana-katta-to
         Mary-non swim-neg-past-comp
         'That Mary did not swim'
         (Sells 1995 4: (4))
      b. [TP [VP [VP [DP Mary-ga] [[V oyog]] [[Neg ana]] [[T katta]] [[C to]]

The tense phrase in Gbe is head-initial. The following example is using the future tense because it is one of the tenses where the tense marker is not inflected onto part of the verb. In example (33), the future marker proceeds the rest of the tense phrase, the VP xɔ̀ kɛ̀kɛ́, with the exception of the specifier. This supports the head-initial evidence for Gbe. The specifier, as mentioned earlier, is not examined when looking at this form of typology because the focus is on the relationship between the head and its complements, whereas the head's relationship to the specifier may be considered under a different ordering system.

(32) Gbe TP Structure
33)  Gbe
   a. dàwé lɔ̀   ná xɔ̀   kɛ̀kɛ́
      man Det Fut buy bicycle
      'The man will buy a bicycle'
      (Aboh 2004 34:(10))
   b. [TP [DP dàwé lɔ̀] [T' [T ] [VP xɔ̀ kɛ̀kɛ́]]]

AspP: relative order of the aspect[edit]

The aspect phrase describes how an action takes place in relation to a verb, as opposed to tense, which focuses on which expresses time. For example, progressive is an aspect that denotes that the action of the verb is currently taking place. The complement of an aspect head is the verb phrase. The following section will examine the aspect phrase in relation to the head-directionality parameter.

In example (34), the verb phrase occurs after the aspectual head: arrived occurs after has in a linear order.[31] Thus, the aspect phrase contains the head-initial parameter. This is also seen in the corresponding syntax tree, since the aspectual head is on the left of the diagram.

(34) English AspP Structure
34) English
      a. John has arrived
         (Lin 2003 2: (5))
      b. [AspP [Asp has] [VP arrived]]]

German contains the same parameter as English. In example (35), the verb weggegangen occurs after the aspectual head ist.[31] This head-initial construction is exemplified in the corresponding tree diagram below.

(35) German AspP Structure
35) German
      a. Hans ist weggegangen
         Hans is left
         'Hans has left.'
         (Lin 2003 2: (7))
      b. [AspP [Asp ist] [[VP weggegangen]]

As with the other phrase types explored previously, Japanese remains consistently head-final, as shown in their aspect structures. In example (36), the imperfective aspect te occurs after its complement ki.[32] When the imperfect aspect combines with a perfective verb like ki, it results in two different readings: progressive and resultant.

(36) Japanese AspP Structure
36) Japanese
      a. Yosio-wa atarasii shatu-o ki-te-ita.
         Yosio-NOM new shirt-ACC put on/wear-IMPERF-PAST
         'Yoshio was putting on his new shirt.' (in-progress reading)
         'Yoshio was wearing his new shirt.' (resultant state reading)
         (Onozuka 2008 6: (9))
      b. [TP [AspP [VP ki] [[Asp te]] [[T ita]]

In Gbe, the aspect phrase contains a head-initial structure. In example (37), the aspect marker occurs before its complement lɛ́sì nà ɖȕ.[33]

(37) Gbe AspP Structure
37) Gbe
   a. Àsíbà tò lɛ́sì nà ɖȕ
      Asiba-IMPERF rice Prosp eat-NR
     'Asiba is about to eat some rice'
     (Aboh 2004 171: (34a))
   b. [AspP [Asp ] [VP [DP lɛ́sì] [Adv nà] [V ɖȕ]]]

Summary of parameters[edit]

The following table summarizes the head-directionality parameter that was discussed previously.The summary of whether the examples from each language were head-initial or head-final was determined by which structure was more common in the data. As the following table shows, languages might not consistently be head-final or head-initial across the language, complicating the typology of analyzing languages as entirely one category or the other. This has posed problems for linguists, and there are further theories below that attempt to explain the phenomenon. In general though, this is the process used to determine the head directionality of a language: one must find examples for all types of linguistic phrases but also try to challenge those results by finding contradictory ones.

English Japanese German Gbe
Verb Phrase Initial Final Final Initial/Final
Noun Phrase Initial Final Initial Initial/Final
Adjective Phrase Initial/Final Final Initial/Final Initial
Prep/Post-position Phrase Initial Final Initial/Final Initial
Determiner Phrase Initial Final Initial Initial
Complementizer Phrase Initial Final Initial Initial
Tense Phrase Initial Final Initial Initial
Aspect Phrase Initial Final Initial Initial
Summary Head-Initial Head-Final Head-Initial Head-Initial

A point for further discussion is comparing and contrasting the positions of adjuncts and complements in relation to the head. Adjuncts, like the ones in English adjectives, seem more "free" in terms of head-directionality than most complements in a language. However, this is for discussion on the adjunct page.

Theories in language typology[edit]

The parametrical approach began with Lucien Tesnière and his hypothesis that language structures can be either Centrifugal or Centripetal. It was later adopted by Chomsky and the structures are now known as head-Initial and head-Final, respectively. The theory of head parameter is often used to act as counterbalance to the maxims proposed by Joseph H. Greenburg [34] in relevance to his theory of word order:

(38)  a.  VO, Prep, N–Gen, N–Adj
      b.  OV, Post, Gen–N, Adj–N
      (Elordieta 2014 2: (1))

The categories in (38) imply that if a language uses Verb-Object order, then all the proceeding features will apply; and if a language uses Object-Verb order, then its respective proceeding features will apply. However, Elordieta posits that many languages (such as Basque) do not fulfill all of the features proposed by Greenburg in (27), and that Greenburg's theory does not predict the exceptions.[35]

Amongst scholars, there is an ongoing debate over the validity of the head-directionality parameter approach. Some of the main conflicts involve whether or not it should be the surface form which determines the parameter, or the underlying (Deep structure) form. A few of the main arguments are outlined as below, including the Surface True approach (the closest to Greenburg's theory of word-order); the Antisymmetry approach, which argues to disprove a binary parameter; and the statistical approach, which looks at variances in parameter within individual languages such that were found in the above data section.

Surface true (word order) approach[edit]

The Surface True approach applies the theory of head parameter to the surface derivations, or even the Phonetic Form(the order which is pronounced in natural speech) of a language, unlike Kayne's proposed theory of Antisymmetry (discussed below) which focuses on movement altering an underlyingly head-initial structure. This section will touch on two articles relevant to this approach.

Head-parameter can only exist in PF: Richards 2008[edit]

Within his article, Richards approaches some issues which have derived in attempting to determine a language's head-parameter, such as the semantic meaning, and the movement and rules needed to maintain universal one-parameter structures. Richards states that the surface distribution of heads indicates that both head-initial and head-final orders are attested in natural language (independently of how such orders are derived).[36] Examples of surface head-initial and head-final structures can be found in the data analysis above, as all of the examples are analyzed based on the surface form of the sentence. Richards further argues that a head parameter must only reside at PF, as it is unmaintainable in its original form as a structural parameter.[37] Antisymmetry looks at the deeper underlying structure.

Existence of true head-final: Takita 2009[edit]

Takita argues that upon application of diagnostics, Chinese is found to be head-initial, as follows Kayne's Universal theory of Antisymmetry; however, upon those same structure tests, Japanese cannot be head-initial, contrary to Kayne's theory. Takita's paper thus argues that the binary head-directionality exists in Universal Grammar, since the movement needs to go from a supposed head-initial structure to a head-final one in Japanese, which would violate other constraints. It is insinuated that languages with dominant head-final structures in their surface derivations are likely following a head-final parameter rather than mass movement to deviate from a head-initial structure. (For a head-initial/antisymmetry analysis of Japanese, please see Kayne 2003).

Head-directionality in Chinese syntax

Takita follows the analysis derived by Lin[38] which tested certain head-final derivations in Chinese which block movement. Lin argues that this is because those head-final derivations are deeply head-initial, thus movement from within the already-moved complement violates Huang's[39] Condition on Extraction Domain (CED), which prohibits movement from within non-complements.[40] Takita thus applies the same diagnostics Lin uses for Japanese head-final derivations to see if they also block movement.

Lin's[38] analysis provides the diagnostics of Move based on the interaction between wh-adverbial zenmeyang 'how' and Sentence-Final Particle (SFP) le with perfect aspectual meaning, as shown in (39).

40a. Le as head of AspP (Adapted from Takita 2009 44: (6a))
(39)  a.  Zhangsan zenmeyang xiu    che ___?
          Zhangsan how       repair car
          'How does Zhangsan repair the car?'
      b. *Zhangsan zenmeyang xiu    che le?
          Zhangsan how       repair car SFP
          'How did Zhangsan repair the car?'
      (Takita 2009 44: (5))
40b. Movement of complement to specifier prevents movement of zenmeyang. (Adapted from Takita 2009 44: (6b))

Zenmeyang can appear in (39a) since it does not have SFP le, but is found ungrammatical when the sentence does contain le (39b). Takita explains that Lin has proposed this is because le is the head of a head-initial aspect phrase (AspP) as seen in (40a). Then, the vP complement to AspP moves to the specifier position, deriving the surface head-final order (40b).[41]

In Chinese, wh-adverbials have to covertly move to SpecCP (the specifier of the complement phrase), so movement of zenmeyang out of the vP (see (40b)) violates the CED defined above.[41] In (39a), since there is no SFP, vP does not move to the specifier of AspP, thus extraction of zenmeyang is possible.

Takita concludes this section of his paper by stating that if a head-final structure in a language is derived from an underlyingly head-initial one, the CED-effect is expected to be observed when extraction from the relevant domain takes place. [42] Thus, if it can be proved that this does not come into effect in Japanese, than Japanese cannot be head-initial. (For further analysis of Chinese phrase structures and derivations, see Takita 2009)

Diagnostics applied to Japanese

Since surface head-final structures are derived from underlying head-initial structures from the act of moving the complements, further extraction from within the moved complement violates CED.

One of the examples of movement which Takita looks at is that of VP-fronting in Japanese. Grammatically, there is not a significant difference between the sentence without VP-fronting (41a) and the sentence where the VP moves to the matrix clause (41b).[43]

(41)   a. [CP1 Taro-ga [CP2[IP2 Hanako-ga [VP2 hon-o    sute]-sae     sita to] omotteiru]
              Taro-NOM        Hanako-NOM    book-ACC  discard-even  did  C  think
          'Taro thinks [that Hanako [even discarded his books]]'
       (Takita 2009 57: (33a)
       b. [CP1[VP2 Hon-o     sute]-saei   Taro-ga [CP2[IP2 Hanako-ga  sita] to] omotteiru
                  book-ACC discard-even  Taro-NOM       Hanako-NOM  did  C   think
          '[Even discarded his books], Taro thinks [that Hanako did ti]'
       (Takita 2009 57: (33b)

In (41a), the fronted VP precedes the matrix subject, confirming that the VP is located in the matrix clause. If Japanese were underlyingly head-initial, (41b) should not be grammatical because it allows for extraction of an element (VP2) from the moved complement (CP2).[43] For further examples, please see Takita 2009.

Thus, Takita shows that surface head-final structures in Japanese do not block movement, as they do in Chinese.

Takita's concluding remarks

It is concluded that because Japanese does not block movement as shown in previous sections, that it is a genuinely head-final language, and not derived from an underlying, head-initial structure. These results imply that Universal Grammar is equipped with the binary head-directionality, and is not antisymmetric. Takita briefly applies the same tests to Turkish, another seemingly head-final language, and finds similar results.[44]

Antisymmetry theory: Kayne 2011[edit]

Main article: Antisymmetry

In contrast to the claim that languages follow the head-initial/final parameter, Richard S. Kayne proposes that all languages are head-initial. Specifically, he argues that the Specifier-Head-Complement order is universal (Antisymmetry Theory), and that all languages have the same underlying order. Antisymmetry theory denies the existence of directionality parameters. One of the main implications of Chomsky's theory is that for every structure a corresponding mirror-image of that structure should also exist; therefore, if a language uses Verb-Object order, syntactic symmetry predicts the existence of a language that uses Object-Verb. Challenging Chomsky's theory, Richard Kayne proposes Antisymmetry, a theory based on two main cross-linguistic asymmetries: syntactic structures found in natural languages seldom have a corresponding mirror-image existing in another language, and no two languages have ever been found to be the exact mirror-copy of each other. Antisymmetry offers a solution to these cross-linguistic asymmetries: by denying the existence of parameters and proposing a universal syntactic structure. Thus asymmetries can be accounted for in terms of different syntactic movements applied by languages. That is, even apparent symmetric pairs such as Verb-Object in English and Object-Verb in Japanese are underlyingly identical and only differ in the application of syntactic movements altering their initial structure, resulting in different surface structures.


Kayne argues that subscribing to a theory that allows both directionalities (like the English/Japanese Verb-Object/Object-Verb dichotomy) implies an absence of asymmetries. The following are a few linguistic examples examining asymmetries in support of Kayne's theory: hanging topics, agreement, and relative clauses.

  • Hanging topics are noun phrases that are a copy of another pre-existing noun phrase. They are usually attached at the edge of the sentence from which the noun phrase was copied.
(42) Der Hans - ich kenne diesen Kerl seit langem
     the.NOM Hans - I know this guy for a long time
     (Nolda 2004 424: (2))
     Hans - I've known this guy for a long time
(42) shows that in German a noun phrase, like der Hans, is inserted onto the left edge of the sentence as a copy of the reference of diesen Kerl. The fact that no occurrence of hanging topics at the right edge of a sentence has been attested[45] provides evidence in support of Antisymmetry.
  • Number agreement is stronger when the noun phrase precedes the verb (V), as stated in Greenberg's Universal 33. The following are some English examples showing number agreement:
(43) a.  There are books on the table.
     b.  There's books on the table.
     (Kayne 2011 7:(23))
(44) a.  Books are on the table.
     b. *Books is on the table.
Kayne argues that the position of there (a dummy pronoun that acts as a subject but refers to the noun phrase books) is due to a process that moved the noun phrase to the left of the verb; when the verb is followed by the noun, the two must agree in number, but when the noun is following the verb, agreement is not necessary. Examples of agreement asymmetry have also been attested in Italian,[46] where the position of the object noun phrase is attributed to movement (bolding and underlying have been used to identify the agreeing/non-agreeing words):
(45) a. Li        ho        visti                                
        3pl     I.have     seen.m.pl        
        'I have seen them'          
     b. *Ho           visti       loro
        *I.have     seen.m.pl      3pl
        [target meaning: 'I have seen them']
     (Kayne 2011 7:(26))
(46) a. Ho           visto         i       libri
        I.have     seen.m.sg     D.m.pl   book.m.pl 
        'I have seen the books'
     b. *Ho           visti         i       libri
        *I.have     seen.m.pl     D.m.pl   book.m.pl
        [target meaning: 'I have seen the books']
     (Kayne 2011 7:(29))
In (45a) the plural verb visti (seen) grammatically agrees when preceded by a plural pronoun (li - them), but cannot agree when it’s followed by a one (loro - them), as in (45b). Similarly, (46a) shows a plural noun phrase "i libri" (the books) following the singular-voice verb "visto" (seen): the verb-noun phrase number disagreement is grammatical. In fact, (46b) shows that <y>verb-noun phrase number agreement is not possible and ungrammatical (in Italian). The fact that number agreement is required on one side of the verb but not the other is an example of language asymmetry.
  • Relative clauses show three main differences:
1. Prenominal relatives (as opposed to postnominal relatives) generally lack complementizers akin to English that
(47) [Popo maai] di tong-tong ne?
     grandmother buy CLF candy PTCL
     what about the candy that grandmother bought?
     (Comrie 2008 729:(6))
(47) shows a prenominal relative clause of Cantonese that lacks complementizers. In contrast, the English translation for (47) uses that as a complementizer (it must be noted that in this case that is optional, as English allows zero relative pronoun constructions).
2. Prenominal relatives (as opposed to postnominal relatives) usually lack relative pronouns
(48) [raketto-o katta] gakusei
     racket-ACC bought student
     the student [who bought the racket]
     (Ozeki 2011 174:(2a))
In (48) [47] the Japanese version lacks a relative pronoun, while the English version (the translation) must use the relative pronoun who in order to maintain the same meaning of the Japanese sentence.
3. Prenominal relatives (as opposed to postnominal relatives) tend to be non-finite. Example (49) is from Quecha (Conchucos), and shows that the verb in the prenominal relative clause is non-finite (as explained by Courtney, the ASP -nqa- indicates completed, ongoing, or at least begun action, making the Verb non-finite).
(49) [runa maqa-nqa-n waka]-ta eika-rqa
     man hit-ASP-3sg cow -ACC see-PST-1sg
     I saw [the cow that the man hit]
     (Courtney 2011 150:(9b))
Overall, the lack of prenominal counterparts to postnominal relative clauses alone supports the asymmetric point of view of Kayne, as asymmetries like the ones shown here are not predicted by symmetrical syntactic theory

In support of Kayne's movement theory, Tokizaki[48] proposes the following complement movement structure:

(50)  a.  [CP C [IP ... [VP V [PP P [NP N [Genitive Affix Stem]]]]]]
      b.  [CP C [IP ... [VP V [PP P [NP N [Genitive Stem Affix]]]]]]
      c.  [CP C [IP ... [VP V [PP P [NP [Genitive Stem Affix] N]]]]]
      d.  [CP C [IP ... [VP V [PP [NP [Genitive Stem Affix] N] P]]]]
      e.  [CP C [IP ... [VP [PP [NP [Genitive Stem Affix] N] P] V]]]
      f.  [CP [IP ... [VP [PP [NP [Genitive Stem Affix] N] P] V]] C]
      (Tokizaki 2011 238:(28))

As per the problem brought up by Elordieta about the application of (50) to Basque, Kayne's theory also poses a problem in that it would involve massive leftward movements within Basque (and other languages') structure to match up to the theorized universal order of an underlying head-initial structure, which would not coincide with a simple grammar ideal.[35]

Kayne offers more examples of linguistic asymmetries, listed below:

Derivational antisymmetry[edit]

This section looks at how Antisymmetry approaches the creation of syntactic strings consisting of at least a subject, a head, and a complement. This process is universally Left-to-Right in Antisymmetry, as it is shown by Kayne in the following two ordered derivations: Head-Complement and Subject-Head-Complement.

Universal <Head-Complement> order[edit]

In order to support the view of an unidirectional underlying structure, probe-goal search is introduced (probe-goal search process is based on the Minimalist program). The idea of probes and goals in syntax is that a head (H) acts as a probe and looks for a goal, a complement (C). The direction of the search is determined by whether the language is probe-initial <Head-Complement> or probe-final <Complement-Head>. Antisymmetry proposes that languages are universally probe-initial, based on statement (51)[49] (51') is a small paraphrasing of the original statement for the sake of clarity):

(51)  Probe-goal search shares the directionality of parsing and of production
      (Kayne 2011 12:(42))
(51') Probe-goal search shares left-to-right direction of parsing and production
      (Kayne 2011 12:(43))

When referring to language comprehension, parsing is the analysis of a string of grammatical constituents (a sentence or phrase). Parsing happens as a hearer hears a sentence: the beginning of a sentence is heard first and the end is heard last. Behaving similarly is production (the encoding of a message into a linguistic form): a speaker produces a sentence in a linear, progressive order, with the beginning of the sentence said first and the end said last. (51) implies that the head (probe) of a sentence must precede the complement (goal), as there cannot be a goal without a probe first; (51') shows that left-to-right can represents the unidirectionality of probe-goal search. According to Kayne, (51) solves the dilemma of choosing between <Head-Complement> or <Complement-Head>, ruling out the existence of <Complement-Head> in favor of always merging head and complement into <Head-Complement>.

Universal <Subject, {Head, Complement}> order[edit]

In order to account for the merge of the subject (S), Kayne proposes the following:

(52) The merger of two phrases is unavailable 
     (Kayne 2011 15:(53))

(52) defines both the <Head-Complement> string and the subject as two distinct phrases if merged externally; to avoid a violation of (53), the subject merges with <Head-Complement> internally, as showed in the following representation:

X-Bar syntactic tree showing (53): S=subject, H=head, C=complement
(53) S H [c...S...]
     (Kayne 2011 15:(53))

(53) shows how the subject, at first internal to the complement (here c to identify the domain of C, which is represented by the brackets), is moved to the left of the head; this avoids the merge of two phrases (in accordance with (52)) by moving a phrase-internal element to a new (and still phrase-internal) position instead. By allowing pair-merge (a process of only merging in pairs - to represent temporal precedence in terms understood as of (53) in the section above), the structure in (54) can be represented as the following:

X-Bar syntactic tree showing (54): S=Subject, H=Head, C=Complement
(54)<S, H> <H, C>
     (Kayne 2011 15:(53))

Notable of the structure in (54) (and the subject movement in (53)) is that it can also be mapped onto X-Bar theory syntactic trees as shown in figure (54) to the right (H', also H or H-Bar, is an intermediate stage of the (head) phrase). Each pair-merge is represented separately from each other, with the subject aligned to the head on its right on one level and the complement aligned to the head on its left on the other (lower) level; this mapping respects the binary branching of X-Bar theory, as well as showing that the head pair-merged to the complement is the same one pair-merged to the subject. Following this process ultimately leads to the precedence relation <Subject, Head, Complement>, which is the sequence that Antisymmetry Theory predicts to be the base structure that every language universally uses (and thus any variation is due syntactic movement).

Statistical approaches to head-placement[edit]

Continuum of head-placement: Tesnière 1967[edit]

In language typology, some scholars argue that there is no absolute head-initial or head-final languages. According to this approach, it is true that some languages have more head-initial or head-final elements than other languages do, but almost all languages contains both head-initial and head-final elements. Therefore, rather than classifying them into fixed categories, languages can be arranged on a continuum with head-initial and head-final as the ends based on the frequency distribution of their dependency directions. Figure (55) on the right illustrates how different languages can be put on a continuum.

Corpus-based statistical analysis of head-placement: Liu 2010[edit]

The evidence of this gradient approach come from a large amount of statistical analysis of the syntactic dependency relation in various of languages, in the light of dependency grammar. Haitao Liu proposed the dependency treebank based method for language typology.[50] He investigated 20 languages using corporas with dependency direction as the typological index. The results showed that some languages are more head-initial or head-final than others. However, no evidence is found for pure head-initial or head- final languages in the sample because each language contains some head-final or head-initial constructions

According to the percentage of head-initial or head-final dependencies, these languages are plotted on a continuum.[51] For instance, referring to the sample languages we analyzed above, Japanese has more head-final elements than other sample languages and therefore is arranged closer to the head-final end of the continuum; English and German, however, have mixed head-initial and head-final dependencies. Therefore, they are plotted in the relatively intermediate position on the continuum. Gbe is somewhere between English/German and Japanese due to many of its phrases possibly having both head-initial and/or head-final constructions.

Correlating lexical inventory with head-placement: Polinsky 2012[edit]

The inconsistency of the head-directionality in language typology (as seen in German verb phrases) can also be investigated by examining the lexical inventories of different languages. Polinsky (2012) proposes that the proportions of verbs and nouns within and across languages may have a correlation with the head-directionality of a given language.[52]

Polinsky investigated the noun-verb ratios across 5 different head-directionality sub-types: rigid head-final, non-rigid head-final, clearly head-initial, SVO/head-initial, and SVO sundry. Basic head-directionality types and examples are shown in the chart below.[53]

Rigid head-final Non-rigid head-final Clearly head-initial SVO/head-initial SVO, sundry
Japanese German Malagasy Indonesian English
Korean Persian Tongan Yucatec Mayan Russian
Tamil Latin most Mayan languages Romance languages
Tsez Irish Bantu languages

After examining the number of verbs and nouns in different languages, the investigated languages can be divided into different head-directionality groups based on the distribution of noun-verb ratios. A strong correlation was found between the head-directionality parameter in a language and the verb-noun ratio in the lexical inventories of the same language. Languages with a scarcity of simple verbs tend to be rigidly head-final, such as Japanese, whereas verb-rich languages tend to be head-initial languages.

See also[edit]


  1. ^ Graffi 2001, pp. 197–198.
  2. ^ Chomsky 1981.
  3. ^ Jackendoff 1977.
  4. ^ Lasnik 2010, p. 41.
  5. ^ Sportiche, Koopman & Stabler 2014.
  6. ^ a b Fukui 1994, p. 4.
  7. ^ Dopke 1998, p. 6.
  8. ^ Aboh 2001, pp. 1–2.
  9. ^ Aboh 2001, p. 2.
  10. ^ Kinyalolo 1992, pp. 1–16.
  11. ^ Smith 1964, p. 6.
  12. ^ Berthold 2005, p. 6.
  13. ^ Naoki 1993, pp. 15–16.
  14. ^ Brousseau & Lumsden 1992, p. 4.
  15. ^ a b Sadler & Arnold 1994, pp. 28–34.
  16. ^ Hinterhölzl 2010, p. 4.
  17. ^ Hinterhölzl 2010, p. 5.
  18. ^ Watanabe 2011, pp. 9–10.
  19. ^ Aboh 2004.
  20. ^ Gillion 1992, p. 15.
  21. ^ Van Riemsdijk 2007, p. 1.
  22. ^ Van Riemsdijk 2007, p. 3.
  23. ^ Vinka 2009, p. 4.
  24. ^ Szabolcsi 1983.
  25. ^ Van Langendonck 1994.
  26. ^ a b Takahashi 2002, p. 2.
  27. ^ Bianchi 1999, p. 251.
  28. ^ a b c d Sells 1995, p. 4.
  29. ^ Kathol 2001, p. 1.
  30. ^ Kratzer 1998, p. 14.
  31. ^ a b Lin 2003, p. 2.
  32. ^ Onozuka 2008, p. 6.
  33. ^ Aboh 2004, p. 171.
  34. ^ Elordieta 2014, p. 2.
  35. ^ a b Elordieta 2014, p. 5.
  36. ^ Richards 2008, p. 280.
  37. ^ Richards 2008, p. 283.
  38. ^ a b Lin 2006.
  39. ^ Huang 1982.
  40. ^ Takita 2009, pp. 42.
  41. ^ a b Takita 2009, pp. 44.
  42. ^ Takita 2009, pp. 45.
  43. ^ a b Takita 2009, pp. 57.
  44. ^ Takita 2009, pp. 59.
  45. ^ Kayne 2011, p. 4.
  46. ^ Kayne 2011, p. 7.
  47. ^ Ozeki 2011, p. 174.
  48. ^ Tokizaki 2011.
  49. ^ Kayne 2011, p. 12.
  50. ^ Liu 2010, pp. 1567–1578.
  51. ^ Liu 2010, p. 1572.
  52. ^ Polinsky 2012, pp. 348-359.
  53. ^ Polinsky 2012, pp. 350.


  • Aboh, Enoch (2001). "Object shift and verb movement in Gbe". Generative Grammar in Geneva 2: 1–13. Retrieved 29 October 2014. 
  • Aboh, Enoch (2004). The morphosyntax of complement-head sequences: Clause structure and word order patterns in kwa. New York: Oxford University Press. 
  • Bianchi, Valentina (1999). Consequences of antisymmetry: Headed relative clauses. Walter de Gruyter. 
  • Broekhuis, H (2013). Syntax of Dutch: Adjectives and Adjective Phrases. Amsterdam: Amsterdam University Press. 
  • Brousseau, Anne-Marie; Lumsden, John (1992). "Nominal Structures in Fongbe". Journal of West African Languages 22 (1): 5–25. Retrieved 29 October 2014. 
  • Chomsky, Noam (1981). Lectures on Government and Binding. Foris Publications. 
  • Comrie, Bernard (2008). "Pronominal relative clauses in verb-object languages". Language and Linguistics 9 (4): 723–733. Retrieved 13 November 2014. 
  • Courtney, Ellen H. (2011). "Learning to produce Quechua relative clauses". Acquisition of Relative Clauses : Processing, typology and function. John Benjamins Publishing Company. pp. 141–172. ISBN 9789027283405. Retrieved 13 November 2014. 
  • Crysmann, Berthold (2005). "Relative Clause Extraposition in German: an efficient and portable implementation". Research on Language and Computation 3 (1): 61–82. 
  • Dopke, Susanne (1998). "Competing language structures: the acquisition of verb placement by bilingual German-English children". Journal of Child Language 22 (3): 555–584. 
  • Elordieta, Arantzazu (2014). Biberauer, T.; Sheehan, M., eds. "On the relevance of the Head Parameter in a mixed OV language". Theoretical Approaches to Disharmonic Word Order (Oxford Scholarship Online). doi:10.1093/acprof:oso/9780199684359.003.0011. 
  • Fukui, Naoki (1993). "Parameters and optionality". Linguistic Inquiry 24 (3): 399–420. Retrieved 26 October 2014. 
  • Gillion, Brendan (1992). "Towards a Common Semantics for English Count and Mass Nouns". Linguistics and Philosophy 15 (6): 597–639. 
  • Graffi, Giorgio (2001). 200 Years of Syntax: A critical survey. John Benjamins Publishing. 
  • Hinterhölzl, Roland (2010). "Collapsing the Head Final Filter and the Head Complement Parameter". Linguistics 20: 35–66. 
  • Hinterhölzl, Roland; Svetlana, Petrova (2009). Information structure and language change: New approaches to word order variation in Germanic. Mouton de Gruyter. 
  • Huang, C-T. J. (1982). "Logical relations in Chinese and the theory of grammar". PhD dissertation, MIT. 
  • Jackendoff, Ray (1977). X syntax : a study of phrase structure. Cambridge: MIT Press. 
  • Kathol, Andreas (2001). "Positional Effects in a Monostratal Grammar of German". Journal of Linguistics 37 (1): 35–66. 
  • Kayne, Richard S. (2003). "Antisymmetry and Japanese". English Linguistics 20: 1–40. 
  • Kayne, Richard S. (2011). "Why Are There No Directionality Parameters?". Proceedings of the 28th West Coast Conference on Formal Linguistics. Somerville: Cascadilla Press. pp. 1–23. Retrieved 13 November 2014. 
  • Kinyalolo, Kasangati (1992). "A note on word order in the progressive and prospective in Fon". Journal of West African Languages 22 (1): 37–51. Retrieved 29 October 2014. 
  • Kratzer, Angelika. "More structural analogies between pronouns and tenses". Proceedings of SALT 8 (92-110). 
  • Kroch, Anthony. "Morphosyntactic Variation". Proceedings of the Thirtieth Annual Meeting of the Chicago Linguistics Society 2: 180–210. 
  • Lasnik, Howard (2010). "Government-binding/principles and parameters theory". Wiley interdisciplinary reviews. Cognitive science 1 (1): 40–50. doi:10.1002/wcs.35. 
  • Lin, Chienjer (2003). "Aspect is result: Mandarin resultative constructions and aspect incorporation". Proceedings of the West Coast Conference on Linguistics (WECOL). 
  • Lin, Tzong-Hong J. "Complement-to-specifier movement in Mandarin Chinese". Ms., National Tsing Hua University. 
  • Liu, Haitao (2010). "Dependency direction as a means of word-order typology: A method based on dependency treebanks". Lingua (120): 1567–1578. 
  • Nolda, Andreas (2004). Topics Detached to the Left: On ‘Left Dislocation’, ‘Hanging Topic’, and Related Constructions in German. Berlin: ZAS Papers in Linguistics. pp. 423–448. Retrieved 13 November 2014. 
  • Onozuka, Hiromi (2008). "On the Resultative Reading of the Imperfective Aspect in English". Australian Journal of Linguistics 28 (1): 1–16. doi:10.1080/07268600701877465. 
  • Ozeki, Hiromi (2011). "The acquisition of relative clauses in Japanese". Acquisition of Relative Clauses : Processing, typology and function. John Benjamins Publishing Company. pp. 173–196. ISBN 9789027283405. Retrieved 13 November 2014. 
  • Ozeki, Hiromi (2011). "The acquisition of relative clauses in Japanese". Acquisition of Relative Clauses : Processing, typology and function. John Benjamins Publishing Company. pp. 173–196. ISBN 9789027283405. Retrieved 13 November 2014. 
  • Polinsky, Maria (2012). "Headness, again". UCLA Working Papers in Linguistics, Theories of Everything (17): 348–359. 
  • Richards, Marc D. (2008). "Desymmetrization: Parametric variation at the PF-Interface". The Canadian Journal of Linguistics 53 (2-3): 275–300. Retrieved 29 October 2014. 
  • Sadler, Louisa; Arnold, Douglas (1994). "Prenominal adjectives and the phrase/lexical distinction". Journal of Linguistics 30 (1): 187–226. Retrieved 29 October 2014. 
  • Sells, Peter (1995). "Korean and Japanese morphology from a lexical perspective". Linguistic Inquiry 26 (2): 277–325. Retrieved 25 October 2014. 
  • Shirai, Yasuhiro; Kurono, Atsuko (1998). "The acquisition of tense-aspect marking in Japanese as a second language". Language Learning 48 (2): 245–279. Retrieved 26 October 2014. 
  • Dominique Sportiche; Hilda Koopman; Edward Stabler (2014). An Introduction to Syntactic Analysis and Theory. Chichester: Wiley Blackwell. 
  • Szabolcsi, Anna (1982). "The possessor that ran away from home". The Linguistic Review 3 (1): 89–102. 
  • Takahashi, Daiko (2002). "Determiner raising and scope shift". Linguistic Inquiry 33 (4): 575–615. 
  • Takita, Kensuke (2009). "If Chinese is head-initial, Japanese cannot be". Journal of East Asian Linguistics 18 (1): 41–61. JSTOR 40345242. 
  • Tokizaki, Hisao (2011). "The nature of linear information in the morphosyntax-PF interface". English Linguistics 28 (2): 227–257. 
  • Van Langendonck, Willy (1994). "Determiners as heads?". Cognitive Linguistics 5 (1): 243–260. 
  • Van Riemsdijk, Henk (2007). "Case in spatial adpositional phrases: The dative-accusative alternation in German". Pitar Mos: A building with a view. Festschrift for Alexandra Cornilescu: 1–23. 
  • Vinka, Mikael (2009). "The syntax of three Japanese postpositions". Lund Working Papers in Linguistics 40: 229–250. 
  • Watanabe, Akira (2011). "Adjectival inflection and the position of measure phrases". Linguistic Inquiry 42 (3): 490–507.