Only complex nodes are differentiated as to their semantic part-of-speech characteristics.
Autosemantic parts of speech. Lexical units that belong to one of the autosemantic parts of speech usually also belong to the corresponding semantic part of speech, i.e. nouns belong to semantic nouns, adjectives to semantic adjectives, adverbs to semantic adverbs and verbs to semantic verbs. This means that the part-of-speech status of the m-lemma usually corresponds to the semantic part-of-speech status of the t-lemma.
However, there is a difference between the part-of-speech status of the m-lemma and the semantic part-of-speech status of the t-lemma with some types of derivation; this involves the following cases:
Pronouns and numerals. Pronouns and numerals are also represented by complex nodes and are treated as either semantic nouns or semantic adjectives (according to their syntactic behavior). For example, který is taken to be a (syntactic and therefore) semantic noun in sentences like Dům, který koupil,... (=The house, that he bought...) and a (syntactic and) semantic adjective in sentences like Který dům koupil? (=Which house did he buy?).
Differences between the values of the m-lemma and t-lemma of pronouns and numerals follow from the fact that all personal and corresponding possessive pronouns (including the reflexives) are assigned a single t-lemma (namely #PersPron) and also from the way some types of derivation are represented at the tectogrammatical level. This concerns the following derivational types:
-
B: the m-lemmas of relative, indefinite, interrogative, negative and totalizing pronouns correspond to nodes with the t-lemmas of the corresponding relative/interrogative pronouns.
-
A: similarly, also possessive counterparts of relative, indefinite, interrogative, negative and totalizing pronouns are represented by nodes with the t-lemmas of the corresponding relative/interrogative pronouns.
-
A and E: the m-lemmas of all types of definite numerals and adverbs with a numerical meaning correspond (at the tectogrammatical level) to the t-lemmas of the corresponding cardinal numerals.
-
D and E: similarly, the m-lemmas of all types of indefinite numerals and adverbs with a numerical meaning correspond to the t-lemmas of the corresponding cardinal numerals.
Other traditional parts of speech. Other parts of speech are represented by other than complex nodes, therefore, no reference to semantic parts of speech is made. These are nodes for particles and interjections (i.e. mostly atomic nodes: nodetype=atom) and nodes for coordinating conjunctions (these form a separate node type, so called paratactic structure root nodes: nodetype=coap).
Prepositions and subordinating conjunctions are not represented by independent nodes at the tectogrammatical level at all (see Section 17, "Prepositions and subordinating conjunctions"), therefore, it makes no sense to refer to the semantic parts of speech with respect to them.
The relations between the semantic and traditional parts of speech are represented in Fig. 5.1. Arrows in boldface indicate that the relation is "prototypical" (nouns prototypically belong to the class of semantic nouns, adjectives to the class of semantic adjectives etc.), simple (thin) arrows indicate the distribution of pronouns and numerals between the semantic nouns and adjectives and interrupted arrows follow the classification based on derivational relations.
