Conversion in English and its implications for Functional Discourse Grammar
Download 202.86 Kb. Pdf ko'rish
|
Conversion in English and its implicatio
D.G. Velasco / Lingua 119 (2009) 1164–1185
1181 reduction of inflections in both the English nominal and verbal system has brought about the break down of the whole word class system. Vogel assumes a classification of the lexemes of languages based on the binary features [ predicatability] and [ reference]. Thus, Mundari- type languages have no word classes because all lexemes show the values [ pred] and [ ref]. In the case of English, however, the loss of verbal inflectional endings has resulted in the infinitival form and the stem being identical and verbs losing their [+pred] feature. Given that nouns also have a negative feature for predicatability, both categories are now assumed to have similar values. Crucially, Vogel (2000:266) argues that the flexible language Tongan shows no signs of ‘an inflectional paradigm characteristic of a specific group of lexemes’. She goes as far as to suggest that Modern English has two overlapping part-of-speech systems, a flexible one and a rigid one, since the language still retains significant aspects of its inflectional system. 15 Note that this formally distinguishes English from other Germanic languages. As pointed out to me by an anonymous reviewer, the analysis presented here is problematic for German and Dutch which seem to show clear grammatical restrictions in conversion. As Don (2005) shows, German does not allow the creation of denominal verbs from stems ending in other than a consonant sound, hence the impossibility of forms like *kaffee-n or *taxi-en. Dutch also presents phonological restrictions in its derivational morphology, but conversion seems to be less restricted: the forms koffie-en (‘to drink coffee’) and taxi-en (‘to go by taxi’) are indeed possible. This confirms, Don argues, an old motto of Dutch linguistics, the intermediate position of Dutch on a historical cline between German and English. However, rather than conflicting evidence, I take this data to support the clear relation between the gradual loss of inflectional morphology and the increasing productivity of conversion. The Dutch inflectional system is comparably richer than that of English, but certainly less than that of German. Significantly, then, the poorer the inflectional system of the language, the greater the functional flexibility of its lexemes. 16 It should also be noted that there is no incompatibility between lexical flexibility and morphological derivation and both processes may be employed by the same language to extend its basic vocabulary (a similar conclusion is arrived at in Don and van Lier, 2007 ). It seems reasonable to hypothesize, though, that languages displaying productive derivational processes will tend to show rather rigid PoS systems, whereas languages with a flexible lexical inventory will make use of fewer derivational processes. Indeed, Mundari seems to have a not fully flexible system as it can also extend its lexicon through derived lexemes ( Hengeveld and Rijkhoff, 2005 :422): (18) a. dal ‘strike’ ! da-n-al ‘a blow’ b. dub ‘sit’ ! du-n-ub ‘a meeting’ c. ol ‘to write’ ! o-n-ol ‘the writing’ D.G. Velasco / Lingua 119 (2009) 1164–1185 1182 15 Jan Rijkhoff (personal communication) has pointed out to me (see also Rijkhoff, 2008 ) that it is inadequate to assume that a flexible word class (or contentive) is a merger of two or more rigid word classes. He argues that flexible word classes are distinct categories in themselves, and therefore, do not show the properties of rigid word classes (transitivity in the case of verbs, gender and number in the case of nouns). Given that converted verbs and nouns in English do show these properties, we would be dealing with a different phenomenon (morphological derivation, most likely). Although at present I have no clear answer to this objection, I see no problem in assuming that contentives may have evolved from differentiated word classes. It obviously remains to be seen if English lexemes will continue losing features such as transitivity and nominal inflections and evolve into a distinct class of their own. 16 Don (2005) also notes that affixed nouns are not usually subject to verbal conversion. I can only account for this by accepting Vogel’s hypothesis that Modern English may have two overlapping lexicons: affixation would thus be restricted to the rigid lexicon and thus segregated from the flexible one. The difference between Mundari and English is that the latter is much richer in morphologically derivational processes, which means that it does not require resorting to lexeme flexibility that much in order to extend its basic vocabulary or vocabulary function. But, significantly, the basic mechanism underlying the flexibility of lexemes in Mundari and the process of innovative verbal creation/conversion is similar in both languages. Consequently, the problem of directionality discussed above takes a different perspective. The question is no longer whether verbs are converted from nouns or nouns from verbs, or whether items are coined and relisted in the lexicon with a new category label or even categoryless as in Marantz (2001) . What I would suggest is that if a thing-denoting item is used as a predicative unit (and to the extent that the process is accepted by the linguistic community) it loses its categorial status, that is, it is relisted (to use Lieber’s term) as a flexible categoryless lexeme. This means that Bauer’s claim above, then, is basically correct. English may be moving towards a different type of lexicon, a more flexible one. By separating frames from lexemes and introducing the semantic theory sketched in section 3 , the FDG model is adequately equipped to handle this process. 5. Conclusions In this paper, I have argued that conversion in English has its roots in innovative coinage as described in Clark and Clark (1979) and Lieber (2004, 2005) . I have also argued that conversion shows that part of the English vocabulary is getting more flexible ( Bauer, 1983; Vogel, 2000 ), possibly due to the loss of inflectional morphology in the recent history of the language. The process underlying conversion is thus assumed to be basically similar to flexible part-of-speech systems in languages such as Mundari. In those languages, lexemes are semantically indeterminate and acquire full meaning potential once they are inserted in a syntactic slot. This is strikingly similar to innovative coinage, as illustrated in the case of conversion from eponyms. The consequences for Functional Discourse Grammar are the following. First, the postulated separation of lexemes and frames finds further support in this paper. In Garcı´a Velasco and Hengeveld (2002) this proposal was applied to transitivity alternations in English (see section 2 ) in which the same lexeme could be used in two different syntactic slots without categorial shift. This paper shows that it is also adequate to account for the multifunctional use of a significant fraction of the English lexicon. Secondly, any account of conversion and lexeme flexibility necessarily requires a semantic theory in which the meaning of lexical items can be constructed and/or adapted cooperatively by speaker and addressee and not just retrieved in one go as a static definition. I have shown that this theory is technically compatible with the postulation of frames in the theory and much more desirable methodologically for a functional theory of language. Acknowledgements I am grateful to Jose´ Luis Gonza´lez Escribano, Eva van Lier, Jan Rijkhoff, Gerry Wanders, the contributors to this special issue and two anonymous reviewers for useful comments and discussion on a previous version of this paper. Needless to say, all remaining errors are my sole responsibility. This research has been financed by the Spanish Ministry of Education and Science through grant I+D MEC-04 HUM2004 01810. Download 202.86 Kb. Do'stlaringiz bilan baham: |
Ma'lumotlar bazasi mualliflik huquqi bilan himoyalangan ©fayllar.org 2024
ma'muriyatiga murojaat qiling
ma'muriyatiga murojaat qiling