Scholarly article on topic 'The well-formed condition for Korean noun incorporation'

The well-formed condition for Korean noun incorporation Academic research paper on "Languages and literature"

CC BY-NC-ND
0
0
Share paper
Academic journal
Pacific Science Review
OECD Field of science
Keywords
{"Head movement constraint (HMC)" / "Korean noun incorporation" / "Theme-only constraint" / "Well-formed condition"}

Abstract of research paper on Languages and literature, author of scientific article — Man-Seob Yang

Abstract In this paper, I aim to discuss the Korean noun incorporation phenomenon. I demonstrate that Korean noun incorporation occurs at the D-structure, syntactically obeying the Head Movement Constraint (HMC) of Travis [14] and Chomsky[7] and Baker [6] and semantically obeying the Theme-Only Constraint (TOC). For the discussion, I first identify the structure of ‘hæ-tot-i’ sunrise-type words, thereby showing that before derivation occurs by adding the nominalizing affix ‘-i’, compounding ‘hæ’ sun and ‘tot-’ to rise should first occur syntactically. Based on empirical data, this argument is strong enough to convince us that noun incorporation is a syntactic word formation process. Based on the syntactic structure of the noun incorporation identified, I show that every noun that goes through syntactic noun incorporation should have the thematic role of ‘theme’, which has been captured as the Theme-Only Constraint. Thus, the syntactic HMC and semantic TOC form a general well-formed condition to effect optimal noun incorporation in Korean.

Academic research paper on topic "The well-formed condition for Korean noun incorporation"

+ MODEL

ELSEVIER

Available online at www.sciencedirect.com

ScienceDirect

Pacific Science Review xx (2015) 1—8

www.elsevier.com/locate/pscr

The well-formed condition for Korean noun incorporation

Man-seob Yang

Daejin University, Republic of Korea Available online ■ ■ ■

Abstract

In this paper, I aim to discuss the Korean noun incorporation phenomenon. I demonstrate that Korean noun incorporation occurs at the D-structure, syntactically obeying the Head Movement Constraint (HMC) of Travis [14] and Chomsky[7] and Baker [6] and semantically obeying the Theme-Only Constraint (TOC). For the discussion, I first identify the structure of 'h«-tot-i' sunrise-type words, thereby showing that before derivation occurs by adding the nominalizing affix '-i', compounding 'h«' sun and 'tot-' to rise should first occur syntactically. Based on empirical data, this argument is strong enough to convince us that noun incorporation is a syntactic word formation process. Based on the syntactic structure of the noun incorporation identified, I show that every noun that goes through syntactic noun incorporation should have the thematic role of 'theme', which has been captured as the Theme-Only Constraint. Thus, the syntactic HMC and semantic TOC form a general well-formed condition to effect optimal noun incorporation in Korean. Copyright © 2014, Far Eastern Federal University, Kangnam University, Dalian University of Technology, Kokushikan University. Production and Hosting by Elsevier B.V. All rights reserved.

Keywords: Head movement constraint (HMC); Korean noun incorporation; Theme-only constraint; Well-formed condition

Introduction

Noun incorporation is unknown in Indo-European languages. In English, there are many Object + Verb compounds (e.g., 'birdwatching', 'babysitting', and 'fingerprinting'). However, the Object + Verb structure is not the case of the verb incorporating the object of the sentence. When these verbs are used transitively, they take other objects: "I have to babysit my aunt's kid" or "I'll fingerprint the suspect." When the verbs do not have an internal argument, they are intransitive: "When I go birdwatching, all I see are mosquitoes". In other words,

E-mail address: msyang@daejin.ac.kr.

Peer review under responsibility of Far Eastern Federal University, Kangnam University, Dalian University of Technology, Kokushikan University.

'birdwatching' does not mean that you necessarily see the 'bird' incorporating into the verb.

Unlike English, Korean has many incorporated nouns with the structure [N + VSTEM + AN1(-i)].The following are some examples of such words.

hœ-tot-i b. mul-pat-i

sun-rise-AN water-receive-AN

'sunrise' '(a) gutter'

haru-sal-i d. kutu-tak-i

a day-live-AN shoe-shine-AN

'(a) dayfly' '(a) shoe-shine boy'

koki-cap-i f. kamok-sal-i

fish-catch-AN prison-live-AN

'fishing' 'living-in-prison'

http://dx.doi.org/10.1016/j.pscr.2014.08.019

1229-5450/Copyright © 2014, Far Eastern Federal University, Kangnam University, Dalian University of Technology, Kokushikan University. Production and Hosting by Elsevier B.V. All rights reserved.

+ MODEL

M.-s. Yang /Pacific Science Review xx (2015) 1—8

For convenience's sake, we will call the words above 'hs-tot-i' sunrise-type words. Concerning the formation process of such words, there might be two analyses: (1) [[N + Vstem]+An(-i)] and (2) [N+ [VSTEM + AN(-i)]]. In this paper, I will first show which of these analyses is more plausible, and then, based on the confirmed word formation process of 'hs-tot-i' sunrise-type words, I will discuss the main concern of this paper, which will be termed the well-formed condition of Korean noun incorporation. It will be shown that the well-formed condition consists of syntactic and semantic constraints that restrict the process of noun incorporation.

Incorporation structure and constraints

As illustrated above, the possible structures for 'hs-tot-i' sunrise-type words are as follows:

bigger compound. As shown below, (3) and (4) provide some more examples of this usage.

[[N + Vstem]+An(-í)] [N + [Vstem + An(-í)]]

(2a) shows that the verb-noun compound occurs prior to the addition of the nominalizing affix AN(-i), while in (2b), the compound occurs after the derivation of [VSTEM + AN(-i)]. The reason that both structures have been controversially assumed for 'hs-tot-i' sunrise-type words is that neither the [N + VSTEM]V structure of 'hs-tot-i' sunrise-type words (e.g., 'hs-tot', 'mul-pat', and 'haru-sal') in (1) nor the [VSTEM + AN(-i)]N structure of sunrise-type words (e.g., 'tot-i', 'sal-i', '^ap-i', and 'pat-i') is used as an independent word2 in modern Korean. For this reason, some have assumed (2a), while others have assumed (2b). In this paper, I will argue that the structure in (2a) is correct. To make a case for the structure in (2a), we will consider the problems with (2b).

Some linguists accept the (2b) structure for the following reasons. First, even though it is true that there is no compound with the [N + V]V structure (e.g., 'hs-tot') or with the [V + AN(-i)]N structure (e.g., [sal + -i]N) in Korean, nominals with the (2b) structure are observed more frequently than those with the (2a) structure when they combine with other nominals in word formation. In other words, although certain word forms (e.g., 'sal-i' living and '^ap-i' catching) are not used as independent words3, they frequently can combine with other nouns to form a

a. m9stm-sal-i

farmhand-live-AN 'living as a farmhand'

c. ok-sal-i

prison-live-AN 'living in a prison'

e. haru-sal-i a day-live-AN '(a) dayfly'

b. kwiyag-sal-i exile-live-AN 'living in exile'

d. f9ka-sal-i

wife's home-live-AN 'living in a wife's home'

a. mal-cap-i

mal4-catch-AN 'a person who measures grain by mal'

c. son-cap-i

hand-catch-AN '(a) handle'

b. param-cap-i wind-catch-AN

'(a) cheerleader '(person)'

d. kal-cap-i

sword-catch-AN '(a) swordsman'

We will now consider the characteristics of 'sal-i' living and '^ap-i' catching in (3) and (4), respectively, in more detail.

Lee [9] analyse words in (3) and (4) as compound nouns and considers 'sal-i' living and 'cap-i' catching as nouns. If he is correct, 'sal-i' living and '^ap-i' catching are derived nominals that are produced by adding the nominalizing affix '-i'. In addition, both of these derived nominals should be able to form bigger derivatives or compounds by combining with other nominalizing affixes or nouns, or they should be able to stand independently, as with the examples below in (5) and (6). However, the data in (7) and (8) show that this capacity to combine or stand alone does not always

[[mol]v+[-i]AN]N chase + AN 'chasing'

[[nol]v+[-i]AN]N play + An 'playing'

[[phul]v+[-i]AN]N solve + An 'solving'

occur.

+ MODEL

M.-s. Yang / Pacific Science Review xx (2015) 1—8

a. [[mol-i]N+[k'un]AN]N

chase-AN + AN('person') '(a) chaser'

b. [[mol-i]N+[p^]AN]N chase-AN + ship '(a) chasing boat'

c. [[phul-i]N+[ran]N]N solve-AN + column '(an) explanation column'

d. [[nol-i]N+[madag]N]N play-AN + field

'(a) playground'

*^ap-i,

a. *sal-i, catch-AN

live-AN

*pat-i, *kel-i,

receive-AN walk-AN

*k'ak-i

b. *^u-i, cut-AN

give-AN

*phal-i *mil-i

sell-AN push-AN

a. *[[sal-i]N + [k'un]AN]N live-AN + AN('person')

c. *[[^ap-i]N + [ran]N]N catch-AN + column

b. *[[sal-i]N + [p^]An]n live-AN + ship

d. *[[^ap-i]N + [madagMN catch-AN + field

Unlike the examples in (5) and (6), some word forms (e.g., 'sal-i' living and 'Cap-i' catching) cannot easily combine with others to form bigger derivatives or compounds, nor can they be used independently, as exemplified in (7) and (8).

Based on the observations above, we can say that in the 'hs-tot-i' sunrise-type words with [N + Vstem + AN(-i)] structures as we see in (1), (7) and (8), 'tot-i' rising cannot be said to be a derived nominal. In other words, the VSTEM and the AN(-i) do not combine to form a derived nominal in 'hs-tot-i' sunrise-type words.

What would it mean to analyse the 'tot-i' rising word type as an affix instead of as a nominal? Such analysis will bring about the following problems. First, in Korean, the 'VSTEM + AffixN(-i)' combination never becomes an affix. Therefore, though the 'VSTEM + AffixN(-i)' combination could be considered an affix through hypercor-rection, the original structure should be [[N + VSTEM]+ AN(-i)], which is represented as (2a).

Second, if the 'tot-i' rising word type, i.e., 'VSTEM + AffixN(-i)', were an affix and were used to show the same distribution in many nominals, then words that belong to such a word type would be considered to have the same meaning wherever they occurred. However, if we consider the nuanced meaning of each word type as it appears in bigger derived nominals, we cannot say that the meaning is the same. We will now take 'sal-i' living and 'cap-i' catching as examples.

a. tyega-sal-i6 :/sal/ / 'to live or stay' wife's home-live-AN

'living in a wife's home'

b. haru-sal-i :/sal/ / 'to manage or extend life'

a day-live-AN '(a) dayfly'

mesum-sal-i :/sal/ / farmhand-live-AN 'living as a farmhand'

to do a role of a farmhand'

a. puk-Cap-i :/Cap/ / 'to grasp with hands'

drum-catch-AN '(a) drummer'

(cf.: AN/-i/here means 'a person who catches something')

b. koki-Cap-i :/Cap/ / 'to catch or capture'

fish-catch-AN 'fishing'

(cf.: AN/-i/here means 'the action of catching something')

c. son-Cap-i :/Cap/ / 'to grasp or catch' hand-grasp-AN

'(a) handle'

(cf.: AN/-i/here means 'something grasped by hands')

As we see in (9) and (10), 'sal-i' living and 'Cap-i' catching, which belong to the same type of word structure [... [Vstem + AN(/-i/) ]], differ in terms of meaning depending to the environments in which they occur. Therefore, based on the data we have discussed above, we cannot treat them as the same affix. Instead, we can explain the meaning differences based on the [[N + Vstem]+-i] structure.

Considering the historical development of the affix [(^ag-i]N,'an artisan', can support this analysis. 'Dag-i' is a diachronically developed affix that has been formed by the addition of '-i', a nominal affix that indicates 'a person' in this case, to the Chinese word 'Cag', which means an artisan, a craftsman or a person who has an excellent skill or technique in certain areas, such as art, handicrafts, etc.Because 'Cag-i' is not an independent word in

+ MODEL

4 M.-s. Yang / Pacific Science Review xx (2015) 1—8

modern Korean, as with 'sal-i' living and 'cap-i' catching, and 'flil-(^ag' a painter/a lacquerer has its full meaning even without the nominal affix '-i', 'tjil-(^ag-i' a painter./ a lacquerer has experienced a change of affixation from

[№]n + №ag]Nk + [-i]NominalAffix]N to [[flil]N + t№ag]N + [-i]NominalAffix]N]N.

Syntactic formation of the '[hx-tot-]' to sunrise-type compounds

In the previous sections, I showed that 'hs-tot-i' sunrise-type words should be understood to have gone through the [[hs-tot]v+[-i]Affix]N formation process. The remaining question will be about the formation procedure of [hs-tot-]v to sunrise, i.e., revealing how and where such verb compound types are formed.

First, I will assume that 'hstot-' to sunrise-type compounds are not lexical compounds but syntactic compounds. One argument for this perspective is that syntactically corresponding sentential—or IP—-structures for 'hstot-' to sunrise-type compounds are easily available in Korean. Another supporting argument is that syntactic affixes, i.e., case suffixes, can be freely inserted between the two morphemes of such verb compounds. Consider the following comparison:

unaccusative verbs, and (11b—e) shows the relationship between the object and transitive verbs.9 Thus, based on the evidence above, we can say that 'hstot-' to sunrise type compounds are a syntactic com-pounds10, not lexical ones.

Syntactic constraint of noun incorporation

We will now consider this paper's main concern. Because the noun 'hs' sun and the verb 'tot' to rise in 'hs-tot-i' sunrise-type words should combine before the nominalizing affix'-i' is attached to the verb and such compounding can be analysed as a syntactic process, there might be Korean noun incorporation processes that can be explained syntactically.11 I assume that a noun, the head of the preceding NP, moves to the head of the following VP to form a syntactic compound in Korean. Refs. [5,6,11] understand this type of head movement as noun incorporation, in which a noun incorporates into a verb. Mithun [11] noun incorporation can be summarized as follows:

a. A noun stem functioning as a subject combines with an intransitive verb to form a new intransitive verb.

b. A noun stem functioning as an object combines with a transitive verb to form a new intransitive verb.

Compounds « Corresponding

a. [hœ-tot-i]N sun-rise-AN 'sunrise'

b. [mul-pat-i]N water-receive-AN '(a) gutter'

c. [kamok-sal-i]N

prison-live-AN 'living-in-the-prison'

d. [koki-^ap-i]N fish-catch-AN 'fishing'

e. [kutu-tak-i]N

shoe-shine-AN (a) shoeshine person'

Sentences

[hs-ka tot-ta]IP sun-NOM rise-DECL 'The sun rises.'

[mul-tl pat-ta]IP water-ACC receive-DECL 'pro receive(s) water.'

kamok-tl sal-ta]IP kamok-ACC live-DECL 'pro serve(s) a prison term.'

[koki-rtl ^ap-ta]IP fish-ACC catch-DECL 'pro catch(es) fish.'

[kutu-rtl tak-ta]IP shoe-ACC shine-DECL 'pro shine(s) shoes.'

We observe that 'hœtot-' to sunrise-type compounds can be easily converted into other bigger syntactic IP structures by inserting case suffixes. Above, (11a) shows the relationship between the subject and

As shown above, in Korean noun incorporation, the preceding noun plays the morphological 'stem' role in combining with the following verb and thereby producing a new compound. (12a) results from combining the noun in the subject position with the following verb, and (12b) results when the noun in the object position moves to the following verb. Under the head-movement hypothesis based on the noun incorporation of (12), we can have the following structures:

a. [[[hs]N]NP[[tot]v]vP-ta120[[[ti]N]NP [hsi + tot]v]vP-ta

'sun' 'rise' DECL

b. [[[mul]N]NP[pat]v]v'-ta0[[[ti]N]NP [muli + pat]v]v'-ta

'water' 'receive' DECL

c. [[[kamok]N]NP[sal]v]v' -ta 0 [[[ti]N]NP [kamoki + sal]v]v'-ta prison' 'live' DECL

d. [[[koki]N]NP[^ap]v]v'ta0 [[[ti]N]NP[kokii+^ap]v]v'-ta 'fish' 'catch' DECL

e. [[[kutu]N]NP[tak]v]v'-ta0 [[[ti]N]NP[kutui+tak]v]v'-ta

'shoe' 'shine' DECL

(13a) indicates that 'hs' sun, which is the head of the NP in the thematic subject position13 under the

+ MODEL

M.-s. Yang / Pacific Science Review xx (2015) 1—8 5

VP, moves to the verb that immediately follows. (13b—e) show that the heads of NPs in the object position (e.g., 'mul' water, 'kamok' prison, 'koki' fish, and 'kutu' shoe) move to the verbs that immediately follow.

Head-movement requires adjacency. In other words, a noun, the head of the NP, and a verb, the head of the VP, should be adjacent to each other to allow the verb to govern its preceding NP-trace. All movements in (13) are licensed; they observe the adjacency condition. In other words, all NP-traces are governed by the following verbs.

When a noun moves to an adjacent verb while observing the adjacency condition, it may produce a left-branching structure in Korean. Consider (14) below:

a. [[hae]N+[tot]v]v

/ \ N V

hae tot

If we accept the unaccusative hypothesis, then we can unify structure (15a) into structure (15b) because in structure (15a), the surface subject is actually assumed to originate as an internal argument of an unaccusative verb, i.e., a theme.

In the (15) structures, the preceding object noun and the following verb are adjacent to one another, and the verb governs the noun. Compare (15) structures with the following (16) structures, in which another head intervenes between the noun and verb:

a. f9lsu-ntn pab-tl mani m9k -ntn -ta. fslsu-NOM rice-ACC much eat -Pres.-DECL 'fslsu eats rice much.'

b. * fslsu-ntn [vp[np ti ][V'[ADVP mani] [pabi + msk]v] V']VP -ntn -ta.

(16) shows that when the adjacency condition is violated, Korean noun incorporation is not allowed. Thus, we can say that the movement for noun incorporation should obey the following Head-Movement Constraint [7,14]:

Based on what we have discussed so far, we can identify the procedure of head movement for Korean noun incorporation as follows: (15)

a. Head Movement from Subject Noun to Verb

/ \ / \ NP V NP V

II I / \

X Y t; Xi V

b. Head Movement from Object Noun to Verb y y

/ \ / \ NP V o NP V

II I / \

X Y ti X; V

Head Movement Constraint (HMC): Movement of a zero-level category b is restricted to the position of a head a that governs the maximal projection t of b, where a q-governs or L-marks t, if a S C.

Semantic constraint of noun incorporation

In this section, I will discuss a semantic constraint that regulates Korean noun incorporation. Though we have revealed that the procedure of Korean noun incorporation is syntactic, some questions still remain: First, is head movement for noun incorporation possible in every sentence with the (15) structure? Second, why does the noun move into the verb? Is the verb acting as a stem in Korean noun incorporation?

The second question is hard to answer. We need additional research and a wider range of empirical data. However, the clue to the first question can be obtained from the thematic relationship between a verb and its preceding noun. In other words, in the noun incorporation structure, a noun is very closely related to one of the possible thematic roles that a verb can play. Consider the following sentences from (18) ((18) repeats (13)):

+ MODEL

6 M.-s. Yang / Pacific Science Review xx (2015) 1—8

(18)14

a. [h^Np [tot]vP -ta 0 Mnp [h^ + tot]v-ta

'sun' 'rise' DECL <Theme>

b. [mul]NP [pat]v -ta 0 [ti]NP [muli + pat]v -ta 'water' 'receive' DECL

<Theme>

c. [kamok]NP [sal]V' -ta 0 [ti]NP [kamoki + sal]v -ta

'prison' 'live' DECL <Theme>

d. [koki]NP [cap]v -ta 0 [ti]NP [kokii + cap]v -ta 'fish' 'catch' DECL

<Theme>

e. [kutu]NP [tak]v -ta [ti]NP [kutui + tak]v -ta

'shoe' 'shine' DECL <Theme>

All nouns in (18) move—i.e., incorporate—into the verbs that follow, and they all have the thematic role of 'theme'.

What would occur when nouns had thematic roles other than 'theme' ? Consider the following instance, in which nouns have thematic roles other than 'theme'

a. Tom-i kaki-ka himdtl-ta

Tom-NOM go-NOM be-difficult-DECL

<Agent>

'It is difficult for Tom to go.'

a'.* [ti] [[Tomi + ka]v + ki]N-ka himdtl-ta

b. totuk-i PO-eke ^ap-hi-dt-ta

thief-NOM PO-'by' catch-Passive-Past-DECL

<Agent>

'A thief was caught by a policeman.' b'. *totuk-i [tj [POi+^ap]v -hi -St -ta.

c. Bill-tn ^ip-e itki-rtl silshan -ta.

Bill-NOM house-'at' staying-Acc dislike-DECL <Locative>

'Bill dislikes staying at home.'

c'. *Bill-tn [ti] [[^ipi + it]v + ki]N-rtl silshan -ta.

d. Bill-tn ^ip-e kaki-rtl silshan -ta. Bill-NOM house-'to' going-Acc dislike -DECL <Goal>

'Bill dislikes going home.'

d'. *Bill-tn [ti] [№ipi + ka]v -ki]N -rtl silshan -ta.

e. nakta-ntn arabia-esS wa-t-ta. camel-NOM Arabia-'from' come-Past-DECL <Source>

'Camels came from Arabia.'

e'. *nakta-ntn [ti] [Arabiai -wa]v -t -ta.

f. Mary-ntn pam-e on-ta

Mary-NOM night-'at' come-DECL <Time>

'Mary comes at night.' f. *Mary-ntn [ti] [pami -on]v -ta

a. Tom-tn nun-i nœri -ki -rtl kitœhan-ta Tom-NOM snow-NOM fall + ing-ACC want-DECL <Theme>

'Tom wants snow to fall.'

a'. Tom-tn [ti] [[nun; + nœri]V -ki]N -rtl kitœhan-ta.

In (19), the thematic roles of adjacent NPs are 'Agent', 'Locative', 'Goal', 'Source' and 'Time'. None of the NP heads in (19) is allowed to move to adjacent vP heads. Though the (19) structures are the same as those of (15) and obey the HMC of (16), noun movement into the following verb position results in, at best, marginal or ungrammatical sentences. On the other hand, in (20), moving the adjacent NP head into the following verb position is allowed as indicated in (18). The difference between the sentences in (18) and (19) is only a difference of thematic roles: [+Theme] in (18) and (20) and [-Theme] in (19). Therefore, we can conclude that noun-to-verb head movement in Korean should obey the following semantic constraint:

Theme-Only Constraint (TOC)

(a semantic constraint of Korean noun incorporation): N can move into an adjacent v for semantic conjunction, if the 0-role of N, a syntactic head, is a theme.

Conclusions

In this paper, I have discussed the internal structure of 'hs-tot-i' sunrise-type compounds, the noun incorporation processes that can be considered as syntactic word formation processes, and syntactic and semantic constraints that regulate Korean noun incorporation processes. To have an optimally noun-incorporated word, the well-formed condition should be obeyed. This condition consists of two sub-constraints: one relating to syntax and the other to semantics. For the syntactic constraint that restricts Korean noun incorporation, I have argued that Korean noun incorporation must obey the Head Movement Constraint (HMC) that [7,14] and [6] have suggested. For the semantic constraint of Korean noun incorporation, I have shown that only nouns with the thematic role of 'theme' can go through the Korean noun incorporation process, which is formally expressed as the Theme-Only Condition (TOC) in (21).

+ MODEL

M.-s. Yang / Pacific Science Review xx (2015) 1—8

1. "An" indicates the nominalizing affix.

2. Unlike words that have the [Vstem + AN(-i)]N structure (e.g., 'tot-i' rising, 'sal-i' living, 'Cap-i' catching, and 'pat-i' receiving), the following [Vstem + AN(-i)]N words ('mol-i' chasing, 'pul-i' solving, and 'nol-i' play(ing)) can appear independently in a sentence. This observation indicates that when these words combine with other nominals to form bigger compounds, the structure will be different from that of 'tot-i', 'sal-i', 'Capi' , and 'pat-i' . A detailed structure is shown below.

(a) [[sastm]N+[[mol]v+[-i]AN]]N deer chase AN 'deer-chasing'

(b) [[munc£]N+[[pul]v+[-i]AN]]N problem solve AN 'problem-solving'

3. Generally speaking, when we say that a word is an independent word, it means that the word is listed in the lexicon. In this paper, I accept this point of view. However, to be more detailed, this point of view should not be perceived as completely plausible, as new words are continuously being formed and are accepted as words through public acknowledgement.

4. A measure containing approximately 18 L.

5. The nominalizing affix [-k'un], a skilful person, can be attached to any nominals in (5) to form bigger derivatives, all of which are acceptable, such as the following:

(a) mol-i-k'un a skilful chaser

(b) pul-i-k'un an excellent problem-solver

(c) nol-i-k'un an excellent cheerleader or amuser who has a great talent for amusing others.

6. All '/-i/'s in (7) seem to express the same meaning of 'an action or behaviour'. On the other hand, the '/-i/'s in (8) all differ from each other in terms of meaning. This fact will support my argumentation. The slightest difference in each verb stem's meaning in (7) and (8) has been identified by The Great Korean Dictionary published by the Keumseong Textbook Publishing Company [10].

7. We can find many Korean words with the '-Cagi' affix.

a. simsul-Cagi: 'a pervasive/contrary/cursed child (person)'

b. yoksim-Cagi: 'a grasping/grabby/greedy fellow'

c. yok-Cagi: 'a foul-mouthed/foul-tongued fellow'

d. s'aum-Cagi: 'a quarrelsome/contentious person'

e. mals'8g~Cagi: 'a troublemaker'

8. In actuality, we have much evidence that confirms the independent usage of 'hstot-' word types in written documents of the Middle Ages. However, it will be beyond of the purpose of this paper to consider the data in more detail. I will thus skip more detailed discussion here.

9. In Korean, syntactic compounds do not involve the relationship between subject-transitive verbs or between object-intransitive verbs.

10 Refs. [1,2] also maintain that NP has the same structure as IP (or TP), which can be observed in the following parallel thematic relationship between NP and its corresponding IP (or TP). Consider the following:

a. Caesar destroyed the city

b. Caesar's destruction of the city

(2) b. DP / \ Spec D' / \ D NP / \ N (XP)

/ \ Spec I' / \ I VP / \ V (XP)

11 Ref. [8] suggests the following four operations for syntactic word formation: (a) head movement and adjunction, (b) merger, (c) fusion, and (d) morpheme fission [11] and [5,6] also explain noun incorporation as a head-movement phenomenon. In other words, they maintain that head movement is for inflectional morphology (e.g., the movement from V to I (or T) and to C) and can also be extended to the movement from N to V.

12. '-ta' is a declarative marker that is located in [INFL, IP]. In actuality, this marker '-ta' does not relate to this section's discussion. Thus, I will just ignore this marker here.

13. Noun incorporation should occur before any other movement or before case assignment. Alternatively, we have to assume another device/procedure for

+ MODEL

8 M.-s. Yang /Pacific Science Review xx (2015) 1—8

syntactically realized case elimination. Here I assume the vP-internal Subject Hypothesis. 14. Recall that all noun-incorporation movements must occur between the D- and S-structures.

References

[1] S. Abney, Functional Elements and Licensing, 1986. Paper Presented to GLOW, Genova.

[2] S. Abney, The English Noun Phrase in its Sentential Aspect, MIT, 1987 (Ph.D. Dissertation).

[5] M. Baker, The mirror principle and morpho-syntactic explanation, Linguist. Inq. 16 (1985).

[6] M. Baker, Incorporation: A Theory of Grammatical Function Changing, University of Chicago Press, Chicago, 1988.

[7] N. Chomsky, Barriers, The MIT Press, Cambridge, Mass, 1986.

[8] M. Halle, A. Marantz, Distributed morphology and the pieces of inflection, in: K. Hale, S.J. Keyser (Eds.), The view from Building 20: Essays in Linguistics in Honor of Sylvian Brom-berger, MIT Press, Cambridge, Mass., 1993.

[9] I.S. Lee, The IC analysis of Korean compound nouns, Korean Linguistics Literature 30 (1965).

[10] K.T.P.C, The Great Korean Dictionary, Published by Keum-seong Textbook Publishing Company, Seoul, Korea, 1994.

[11] N. Mithun, The evolution of noun incorporation, Language 60 (1984) 845-895.

[14] L. Travis, Parameters and Effects of Word Order variation, MIT, Cambridge, Mass, 1984 (Ph.D. Dissertation).