A Minimalist Note On Korean Honorific Agreement

Abstract

Honorification has been a major concern to any serious linguistic research into the Korean language. In this presentation, first, we will briefly review major works on the phenomenon, which will reveal some "major" issues as in (1): (1) a. Is Korean honorification a grammatical agreement phenomenon or a pragmatic one? b. Does honorification correlate with a certain sentential constituent, e.g. the subject? c. What is the best way to describe this phenomenon? (i) What is the correct picture of the distribution of the honorific suffix si? (ii) Does Korean grammar have an agreement-functional category Agr? (iii) What are the correct mechanisms for this phenomenon? Second, we will review the basic properties and the standard treatments of grammatical agreement in other languages like English (Chomsky 2000, 2001, among others): (i) asymmetry between the trigger (or Goal) and the target (or Probe) captured in terms of interpretability of formal features; (ii) locality between the trigger and the target captured by Attract F or Agree; and (iii) syntactic visibility or activeness captured by means of presence of uninterpretable features in both parties. We will then review five basic properties of the Korean honorification phenomena, some of which have been used to argue that this linguistic portion is best characterized pragmatically: (i) pragmatics-dependency, (ii) the tripartite division in honorification domain, (iii) gradability, (iv) de-animation (Park 1975), and (v) the so-called experiencer-hood import of honorification (Im 1985). The basic tenet of our approach is that Korean honorification is best analyzed when we recognize both its pragmatic and syntactic aspects. We combine Chang`s (1973) and I. Lee`s (1974) ideas to capture the property of pragmatics-dependency, and propose that HONOR, a three-place honorific operator (or feature), may be assigned to a noun (phrase) and always has the speaker as the value for its first variable. We then propose an interpretation principle that a multiple occurrence of HONOR in a noun phrase intensifies the degree of honorification to account for the property of gradability. As for the remaining three properties enumerated above, then, we suggest that they seem to show that Korean honorification needs syntactic treatment as well: First, the property of the tripartite division in honorification domain and the "spreading" occurring in them are significantly reminiscent of Chomsky`s division of clausal structure in terms of the functional categories v, T, and C. We propose a two-step Agree: The first step involves a valuation of the uninterpretable honorific feature [uH] of the Agr by a [+H] NP ; the second step involves the potentially multiple valuation by the honorifically valuated Agr of [uH] features on nominal or verbal elements in its "scope" (or domain) (cf. Hiraiwa 2000). The property of de-animation, next, may be understood as reflecting the universal property of syntactic visibility. Third and lastly, the semantic import for experiencer-hood may be best understood as a result of the semantic interpretation of the functional projection of the honorific Agr head (Hale and Keyser 1993). Im`s (1985) cases where si-honorification seems to be triggered by a non-subject constituent may be shown either that they are in fact triggered by subjects, or that Agree may capture them, e.g. (2) from (Im 1985). (2) Cwunim-i-si-ye, na-lul kwupe salphi-sose. Lord+H-Cop-H-Voc, I-Acc pay+attention+to-Exh `Dear Lord, please pay attention to me.` The last part of this presentation is devoted to a more syntactic argumentation on the issue of whether Korean grammar has a separate functional category Agr responsible for honorification agreement. I present arguments for and against an independent Infl head and then for and against an independent Agr head, refuting the reactional ones. Crucial pieces of evidence for the Infl head and for the Agr head are the separability of inflectional morphemes from the verb stem as shown in (3), and the seperability of the honorific morpheme si from both the verbal stem and the tense morpheme as in (4)-(5), respectively: (3)-(4) are from Y. W. Park 1992; (5) from J.-M. Yoon 1990. (3) a. Mother went shopping, and father did so too. b. emeni-kkeyse sicang-ey ka-si-ess-ko apeci-kkeyse-to kuli ha-si-ess-ta mother-NomH market-to go-H-Past-and father-NomH-also so do-H-Pst-Dec `Mother went to the market, and father did so, too.` (4) [apenim-i [[nolay-lul pwulu-]-ko/-myense [chwum-ul chu-]]-si-ess-ta. father.Hon-Nom song-Acc sing-and/-while(?) dance-Acc dance-Hon-Pst-Dec `Father sang a song and danced a dance.` (5) [AgrP apenim-i nolay-lul pwulu-si]-ko [AgrP emenim-i chwum-ul chu-si]-ess-ta. father-NomH song-Acc sing-H-and mother-Noon dance-Acc dance-Hon-Pst-Dec `Father sang a song and mother danced.` Concluding remarks of this presentation will include the following: (i) Some major properties of Korean honorification seems to be better captured with Agree in a Minimalist program-type treatment; and (ii) empirical evidence still favors postulation of -si- and -ess- as syntactically independent functional heads. Remaining problems include: (i) Does this Agr have an interpretable feature like other functional heads T and C? (ii) How similar do Korean and Japanese in honorification (cf. Kuno 1987 and Niinuma 2001, among others)?