Academia.eduAcademia.edu

Figure 3 – uploaded by Mark W. Post

See full PDFdownloadDownload figure
Table 2. Capacity for nominalization by khwaam or kaan by semantic type by khwaam while core verbs are nominalized exclusively by kaan. Distribu- tional overlap is found in the semantic margins of the two classes, among HUMAN PROPENSITY and STATE/ACHIEVEMENT terms. The nominalization test therefore clearly does not reliably discover all and only the members of an adjective or verb class, as cited accounts have already indicated. However, it appears to discover an ordered relationship in the distribution of terms from two semantically-defined classes. Namely, it finds that while terms sharing like semantics pattern together, terms from the relatively dissimilar semantic cores of the two sets also contrast in their distribution.

Table 2 Capacity for nominalization by khwaam or kaan by semantic type by khwaam while core verbs are nominalized exclusively by kaan. Distribu- tional overlap is found in the semantic margins of the two classes, among HUMAN PROPENSITY and STATE/ACHIEVEMENT terms. The nominalization test therefore clearly does not reliably discover all and only the members of an adjective or verb class, as cited accounts have already indicated. However, it appears to discover an ordered relationship in the distribution of terms from two semantically-defined classes. Namely, it finds that while terms sharing like semantics pattern together, terms from the relatively dissimilar semantic cores of the two sets also contrast in their distribution.

Related Figures (7)

Table 1. Properties distinguishing verb sub-types in Lao (Enfield 2007: 270)
Table 1. Properties distinguishing verb sub-types in Lao (Enfield 2007: 270)
of verbs. Enfield’s main subclassification criteria, and the verb class-taxonomy for which they are the primary basis, provide a critical point of reference and are worth reproducing here (Table | and Figure 1).
of verbs. Enfield’s main subclassification criteria, and the verb class-taxonomy for which they are the primary basis, provide a critical point of reference and are worth reproducing here (Table | and Figure 1).
Table 3. Properties of lexical compounds, attributive clauses and relative clauses com- pared To summarize, attributive clauses closely resemble lexical compounds, both formally (in being directly postposed to a nominal) and semantically (in de- noting a class of entity in the cultural world). At the same time, attributive clauses are like relative clauses in being syntactically compositional rather than non-compositional or idiosyncratically compositional, and in implying a head- coreferential “gap”. Accordingly, attributive clauses may be paraphrased via a relative clause construction — when appropriate semantic and pragmatic condi- tions exist — with no change in the basic denotation of the expression. By con- trast, it is not possible to paraphrase most true lexical compounds via a relative clause without changing the denotation; for example, a relative clause para- phrase of (16b), jaa thii baa [medicine REL crazy] forces an Attributee subject reading ‘medicine that is (itself) crazy’, since this is the only type of participant role which may be ascribed to a relativized adjective in Thai. Finally, however, attributive clauses in Thai differ from relative clauses syntactically in permit- ting relativization on attributee subjects only, whether the relativized modifier is an adjective or intransitive/transitive verb; they do not license the set of rel- ativizations on objects and non-core arguments which are licensed by relative clauses. Table 3 summarizes the similarities and differences among the three construction types.
Table 3. Properties of lexical compounds, attributive clauses and relative clauses com- pared To summarize, attributive clauses closely resemble lexical compounds, both formally (in being directly postposed to a nominal) and semantically (in de- noting a class of entity in the cultural world). At the same time, attributive clauses are like relative clauses in being syntactically compositional rather than non-compositional or idiosyncratically compositional, and in implying a head- coreferential “gap”. Accordingly, attributive clauses may be paraphrased via a relative clause construction — when appropriate semantic and pragmatic condi- tions exist — with no change in the basic denotation of the expression. By con- trast, it is not possible to paraphrase most true lexical compounds via a relative clause without changing the denotation; for example, a relative clause para- phrase of (16b), jaa thii baa [medicine REL crazy] forces an Attributee subject reading ‘medicine that is (itself) crazy’, since this is the only type of participant role which may be ascribed to a relativized adjective in Thai. Finally, however, attributive clauses in Thai differ from relative clauses syntactically in permit- ting relativization on attributee subjects only, whether the relativized modifier is an adjective or intransitive/transitive verb; they do not license the set of rel- ativizations on objects and non-core arguments which are licensed by relative clauses. Table 3 summarizes the similarities and differences among the three construction types.
Table 4. Number of non-nominal adnominal modifications, classified according to type. Total A = Total number of mentions by lexical class (ADJ or V); Total B = Total number of mentions by modification type (Direct or Marked)
Table 4. Number of non-nominal adnominal modifications, classified according to type. Total A = Total number of mentions by lexical class (ADJ or V); Total B = Total number of mentions by modification type (Direct or Marked)
The results of the frequency counts are presented in Tables 4 to 6; Table 4 ex- presses the numbers of each type of adnominal modification identified as frac- tions of the total number of adnominal modifications. Tables 5 and 6 express the frequencies of each type of adnominal modification identified as percents of total lexical class type mentions and adnominal modification type mentions, respectively. Table 5. Frequency of non-nominal adnominal modification types, expressed as percents of total lexical class mentions
The results of the frequency counts are presented in Tables 4 to 6; Table 4 ex- presses the numbers of each type of adnominal modification identified as frac- tions of the total number of adnominal modifications. Tables 5 and 6 express the frequencies of each type of adnominal modification identified as percents of total lexical class type mentions and adnominal modification type mentions, respectively. Table 5. Frequency of non-nominal adnominal modification types, expressed as percents of total lexical class mentions
Table 6. Frequency of non-nominal adnominal modification types, expressed as percents of total modification type mentions noun phrases with temporal or procedural content (such as in (20b)).7° Rel- ative clauses with an adjectival nucleus were extremely rare, with only three mentions across all eight texts. Notably, in all three cases the “adjectives” in question were either tem ‘full’ or waay ‘empty’, both of which are perhaps only marginally to be considered as adjectives.7!
Table 6. Frequency of non-nominal adnominal modification types, expressed as percents of total modification type mentions noun phrases with temporal or procedural content (such as in (20b)).7° Rel- ative clauses with an adjectival nucleus were extremely rare, with only three mentions across all eight texts. Notably, in all three cases the “adjectives” in question were either tem ‘full’ or waay ‘empty’, both of which are perhaps only marginally to be considered as adjectives.7!
Table 7. Summary comparison of adjectival and verbal distributions in selected Thai constructions ity judgements in this area would appear high. For example, sdot ‘single (not in a relationship)’ is attested as both intransitive predicate and copula com- plement and can be nominalized in khwaam just like a verblike property term (41).
Table 7. Summary comparison of adjectival and verbal distributions in selected Thai constructions ity judgements in this area would appear high. For example, sdot ‘single (not in a relationship)’ is attested as both intransitive predicate and copula com- plement and can be nominalized in khwaam just like a verblike property term (41).
Related topics:
Languages and LinguisticsGrammatical Categories and CognitionThai Language
580 California St., Suite 400
San Francisco, CA, 94104
© 2025 Academia. All rights reserved