www.pearsoned.ca/text/ogrady/syntax/variation Variation in phrase structure Even where languages have the same categories, there can be variation in terms of how they are assembled. Part of this variation can be traced to a head position parameter that offers two versions of the X' schema one in which the head precedes its complement and one in which the head follows its complement as shown in figure 1. English adopts the head-initial option, as shown by the fact that verbs, prepositions, and other heads precede their complements (as in watch TV, in school, and so on). In Japanese, in contrast, heads consistently occur in the final position within their phrase (see figure 2). Thus, the noun comes at the end of the NP, the verb at the end of the VP, and so on. Because Ps occur at the end of the PP, they are called postpositions rather than prepositions.

As summarized in table 1, these word order differences reflect the positioning of heads with respect to their complements not the presence of an entirely new type of syntactic system. Variation in whether Move is used to form particular sentences As we have already seen, English makes use of the Move operation to form both yes-no questions (which involve Inversion) and wh questions (which involve Wh Movement as well as Inversion). However, this is not the only way that questions of this sort are expressed.

For example, in Tamil (a language of India), yes-no questions are signaled by the presence of the morpheme -ā at the end of the sentence, rather than by an Inversion transformation. (The diacritic indicates a dental point of articulation; the diacritic marks a long vowel; l is a retroflex liquid.) As depicted in figure 3, corresponding to sentence 1b), Tamil adopts the head-final version of the X' schema. Thus, the V comes at the end of the VP, I at the end of IP, and C (containing the question morpheme) at the end of CP. Just as some languages form yes-no questions without the help of the Inversion transformation, so some languages form wh questions without using Wh Movement. Languages of this type include Japanese, Korean, Tamil, Chinese, and Thai. The following example is from Thai.

2) Khun ʔaan ʔaray? you read what What did you read? It seems that the Q feature in languages such as Thai is not strong enough to attract wh words, with the result that wh questions are formed without the help of the Move operation (see table 2). Verb-initial languages Languages with verb-subject-object word order provide a special challenge for syntactic analysis, ultimately revealing yet another way in which languages differ from each other. The following example is from Welsh. 3) Lladdodd y ddraig y dyn killed the dragon the man The dragon killed the man. At first glance, the word order here is puzzling, since neither version of the X' schema discussed above allows a subject to occur between the verb and its complement.

In order to accommodate languages of this type, it has been necessary to consider a radical idea namely that the subject originates in the specifier position of VP in all languages (see figure 4). (If this is right, preverbal adverbs have to be reanalyzed as something other than specifiers perhaps modifiers.) In a verb-medial language such as English, the subject raises from the specifier position within VP to the specifier position of IP in surface structure. In a verb-initial language such as Welsh, by contrast, the subject remains in the specifier position within VP, but the verb raises to the I position (just as it does in French). As shown in figure 5 and table 3, this gives the correct final order for both types of language.

This analysis has two immediate advantages. First, it allows us to extend the X' schema to verb-initial languages, which is clearly desirable since the schema is supposed to be part of Universal Grammar. Second, it allows us to account for the differences between verb-medial and verb-initial languages in terms of the application of Move operations that are independently attested in human language.

Exercises 1. The following data is from Korean. You may ignore the nominative (subject) and accusative (direct object) markers for the purposes of this question. a) Terry-ka ku yeca-lul coahanta. Terry-Nom that girl-acc likes Terry likes that girl. b) I noin-i hakkyo ey kassta. this man-nom school to went This man went to school. c) Sue-ka chinkwu eykey chayk-ul ilkessta. Sue-NOM friend to book-acc read Sue read the book to a friend. i) Draw the tree structure for each of the Korean sentences. ii) Which version of the X' schema does Korean use? 2. Consider the following Selayarese data. a) Laʔallei doeʔ injo ibaso. took money the Baso Baso took the money. b) nraʔbai sapon injo. collapsed house the The house collapsed. c) Lataroi doeʔ injo ri lamari injo ibaso. put money the in cupboard the Baso Baso put the money in the cupboard. i) Draw a tree structure for each of these sentences. ii) How does the X' schema for Selayarese differ from the one for English?