This chapter examines the relation between the structure of words as linguistic objects and their conceptual content. It addresses two questions: what are the primitives of lexical semantic interpretation, and how they are expressed in the grammatical and morphological representation of a lexical item. The answer involves a characterization of roots as theoretical objects, followed by an argument to the effect that it is not roots, but larger structures of variable size which relate to lexical concepts. An in-depth discussion of nouns leads to the claim that the conceptual content of a lexical item does not reflect its grammatical structure, because a concept is not the meaning of a linguistically defined unit, but a language-external cognitive content, globally associated with the lexical word as a whole.
In diachronic lexicology, word formation (henceforth: WF) is generally treated as one of three strictly distinct types of processes of lexical innovation, the two others being meaning change (MCh) and borrowing. A unitary theoretical framework based on Construction Grammer and integrating different types of lexical innovation will show that, from an onomasiological point of view, WF and MCh in particular are only two ways of switching from a construction 1 expressing a concept C1 to a formally and semantically related construction 2 expressing a concept C2. From this perspective it is not surprising to find multiple combination, intertwining and interaction between MCh and WF: WF and MCh as subsequent steps in chains of diachronic change; WF and MCh as designative alternatives in lexical innovation; WF and MCh as components of the phenomena of lexical ellipsis and folk-etymology; MCh governing the diachronic evolution of already existing WF devices; MCh creating new WF devices.
Morphologically and syntactically conditioned alternations indirectly code their conditioner. The distinction between semantic information that is coded by a property of an expression and semantic information that conditions and constrains a property of an expression brings rigor into linguistic description and makes us understand an important mechanism of interpretation and reanalysis of linguistic structure. The paper provides a theoretical basis for describing indirect coding both on the paradigmatic and on the syntagmatic axis, but then focuses on syntagmatically mediated coding.
In this paper we discuss two cases of seemingly polysemous derivational affixes: the Ewe suffix -ví, originally a noun meaning “child”, which has acquired a number of different semantic values in word formation, as e.g. “inexperienced” (núfíáláví “inexperienced teacher”) or “person who adheres to the typical behaviour of a group” (amredzóví “someone who behaves like a foreigner”, and the Mandarin Chinese suffix -bā, originally a lexical morph meaning “bar”, which also helps to form locative nouns conveying a broad range of meanings, as e.g. in yóuxìbā (game-bā) “amusement arcade” or in yǎnbā (eye-bā), a kind of optometry clinic. We shall show that apparent polysemy may be a consequence of generalization, rather than of specialization in meaning, and that the mechanisms involved in the evolution of derivational affixes are mostly analogous to those of grammaticalisation.
The paper investigates the semantics of non-suffixed adjectival adjective-noun (AN) compounds in Slavic languages (especially in Slovak and Czech) within a lexeme-based approach in morphology (Aronoff 1976, 1994; Fradin 2003). Two types of AN compounds exist in Slavic languages: a suffixed one, where the suffix has scope over the whole complex base (such as vysok-o-škol-ský), and a “suffixless” one, where the lexical category is marked by means of an inflectional ending (such as modr-o-ok-ý corresponding to the well-known type of compounds like blue-eyed). The article shows, on the basis of the data drawn from large corpora, that there is a basic semantic difference between the two types and it attempts at formulating the principal constraint involved in the formation of the suffixless compounds. The paper puts forward a Principle of Integrated Meronymy as the basic semantic criterion that regulates the possibility of formation of this type of adjectival compounds.
Emerging evidence suggests that integrating the constituents of compound words involves semantic composition and that this meaning construction process draws on relation information linking the constituents. Research with novel compounds (for which semantic composition is obligatory) has found that relation structures compete for selection during semantic composition and that increased competition results in increased processing difficulty. The current project investigates whether relation competition occurs in the processing of established transparent and opaque English compounds. The results indicate that more relation competition is associated with more difficult processing of compound words, even those that are semantically opaque. This indicates that a relation-based semantic composition process is initiated during the processing of established compounds, even for semantically opaque compounds where the final interpretation cannot be relational. Understanding the semantic composition process is critically important in creating a complete theory of compound processing.
The abessive suffix of the Mordvin languages (Erzya and Moksha) expresses the meaning “without”. Diachronically, the suffix presents an example of a process of grammaticalization in which a derivational ending has developed into a case suffix. However, the process has not been completed, as the derivational meaning has also been maintained in both Erzya and Moksha. Consequently, in contemporary Mordvin languages the abessive form has a twofold character: it can be used both adverbially (as a case form) and adnominally (as an adjective derivation). This article investigates both the derivational and inflectional aspects of the abessive in the Mordvin languages and compares its uses with some other cases.
Most word-formation theories assume that derivatives in general tend towards canonicity, i.e. one-to-one correspondence between form and meaning and thus full morphosemantic transparency. However, form-meaning mismatches actually are widespread, both in terms of languages and word-formation rules. These mainly fall into two types: over-marking and under-marking. In this paper we propose a classification of these deviations, distinguishing between a derivational and a lexical level. We illustrate this classification with examples from French and other languages (English, Italian and Dutch). We sketch a unified analysis of these deviations within a word-based framework. We propose to analyse the relative importance of canonicity and discrepancies in word formation from the perspective of the interaction between the speaker and the hearer.
Nominal compounds and corresponding phrases provide a suitable test bed for inquiry into the interface and demarcations between syntax and morphology. In this article we show that the distinction between the two constructions is blurred from a theoretical perspective and investigate processing and cognitive aspects. After an examination of structural as well as semantic properties of the two types, we report on three studies (memorization, questionnaire, and reading time) that experimentally juxtapose German adjective-noun constructions of phrasal and morphological kinds. The results of these studies indicate cognitive differences between compounds and phrases. Considering alternative explanations, we argue that our data supports the view of two separated structural levels of grammar.
We report on a psycholinguistic study of semantic transparency among English compound words. We employed the P3 technique (Libben et al. 2012), which combines elements of three experimental paradigms: progressive demasking, naming, and word typing. Both the progressive demasking data and the word typing data showed a graded effect of semantic transparency associated with the number and location of semantically opaque constituents in the compound. Typing duration effects were evident at the constituent boundary, supporting observations first made by Sahel et al. (2008). We suggest that these data point to the value of letter typing durations in the analysis of morphological processing and the importance of a psychocentric perspective on lexical processing – one that emphasizes the psychological nature of morphological structures.
This paper examines the interaction between form and meaning in creole word-formation, drawing on evidence from Kriyol, a Portuguese-based creole spoken in Guiné-Bissau. The goal will be to illustrate how full reduplication interacts in complex ways with other morphological operations such as conversion, derivation and inflection. A formal morphological analysis of the layering between form and meaning will be sketched, within Construction Morphology, which treats full reduplication as a genuine lexeme-formation process. Within this word-based theory, the interaction between morphological operations is captured through the unification of construction schemas.
Compounds such as actress-cum-model, pub-cum-hotel or military-cum-religious represent mainly an alternative construction to coordinative compounds of the singer-songwriter type. Although most of them are double-headed, some can show semantic left-headedness. We have detected six main semantic groups out of a corpus of 300 types of -cum- nominal compounds. They correspond, in descending order of frequency, to people’s professional or leisure activities; buildings, rooms, and other places; literary and audio-visual genres; pieces of furniture, clothes, and other objects; abstract nouns; and toponyms, and anthroponyms. Under the heading “People”, we consider two main subheadings: team-sports players and antonymous occupations or functions.
This contribution studies the synchrony and diachrony of German plural variants. Plural doublets in the sense of overabundance must be restricted to variants with the same meaning used by one and the same speech community. But there is a preference for distinctiveness which assigns different pragmatic or semantic meanings to variants. This includes also a preference for iconic relationships between form and meaning. The assumption of a preference for biuniqueness appears to allow greater generalizations than the ideal of canonical paradigm forms.
Noun Incorporation Constructions commonly involve one noun root and one verb root, but in some languages the verbal member may be compounded with two nominal roots, which almost always have the semantics of a theme and a non-theme (a problematic fact for Baker’s 1988 theory). Moreover, if the non-theme noun specifies instrument or manner, the linear order in which the two nouns appear may vary, one noun taking scope over the other. This fact too can be a problem for Baker’s theory, as well as for any theory making use of a fixed Thematic Hierarchy. By reviewing the most salient features of Noun Incorporation Constructions in general, a theoretical apparatus is proposed which dispenses with the notion of head movement in favor of XP movement, yet preserves the idea of a Thematic Hierarchy by means of a phase-based model of word formation.
The central claim of this paper is that, besides phonological optimization, a major function of shortening words by different techniques such as clipping or acronymy is a controlled reduction of morphosemantic transparency in lexical domains, where partly or entirely opaque words are preferred to completely transparent ones. E.g., the functions of uniquely identifying and individualizing a referent by a proper name are better fulfilled by opaque labeling than by a transparent descriptive nomination. In fact, a study of shortening techniques used for official and commercial proper names contrasted with non-onymic words from the general lexicon and certain jargons in German, Farsi and Standard Chinese, three languages extremely different with respect to linguistic type and writing system, reveals a clear contrast between highly transparent shortenings preferred in the non-onymic lexicon versus less transparent or even opaque ones in the proper names, although the preferred shortening techniques themselves vary from language to language.
The paper concerns the structural factor of ‘root transparency’ in relation to the morphology-meaning interface in the mental lexicon of Hebrew. Semitic root structure was analyzed by comparing responses of native speakers of Hebrew on a set of written tasks. Target items included both nouns derived from transparent full triconsonantal roots (e.g. migdal “tower” from the root g-d-l) and nouns derived from more opaque defective biconsonantal roots (e.g. taxazit “forecast” from the root x-z-y). Responses to items based on full roots demonstrated that they were across-the-board more accessible and easier to process than their defective counterparts. Moreover, the difference between performance on items derived from full compared with defective roots was found to interface with phonology, semantics, and syntax, as well as with familiarity/frequency. These results indicate that root transparency plays an important and pervasive role in the mental lexicon of Hebrew.
This paper will discuss several lexical semantic differences between mimetic verbs and conventional prosaic verbs in Japanese, and show that meaning for mimetic verbs calls for a treatment distinct from that of lexical verbs. Taking into account the notion of ‘affect-imagistic dimension’ (Kita 1997) in which a mimetic base is represented in terms of its images, I will argue that an analysis that incorporates the basic premise of frame semantics and construction grammar can accommodate the seemingly idiosyncratic and unconventional lexical semantic properties of mimetic verbs. The wide range of semantic frames of mimetic verbs is thus fed by the numerous images of the mimetic base and the specific constructions in which they appear.
The term ‘deponent’ traditionally refers to a class of verbs in Latin or Greek which is characterized by a mismatch between their morphological form and their syntactic and semantic context. Deponent verbs appear in passive form but in active syntax/semantics. In this paper, I argue that this mismatch between form and function is not as uncommon as one might expect. In particular, I show that we find the same kind of mismatch with unaccusative verbs in many languages amongst which are English or the Romance languages. The argumentation will be based on a detailed investigation of how deponent and unaccusative verbs behave in all modules of the grammar. Based on the findings of this discussion, I propose a unified morphosyntactic analysis for deponents and unaccusatives. The analysis elegantly captures the particular behavior of these verbs by invoking two principles: Identity Avoidance and Lexical Override.
The present paper discusses deponent verbs in Modern Greek. Deponents are claimed to be idiosyncratic, non-canonical verbs representing a mismatch between morphology and syntax/semantics, as they are intransitive verbs exhibiting a morphological form, non-active, without there being an active transitive counterpart. Since non-active morphology is standardly associated with de-transitivization, deponents are taken to be exceptional in this respect. We compiled a deponent-verb corpus of Modern Greek and examined its morphological, syntactic and semantic structure. The results of this study revealed that most deponent predicates are actually reflexives/reciprocals, anticausatives and passives; thus, they instantiate verbal alternations which typically surface with non-active morphological marking. For this reason, we conclude that the morphology of deponent verbs in Modern Greek has, contrary to their traditional definition, a canonical function.
This chapter examines the relation between the structure of words as linguistic objects and their conceptual content. It addresses two questions: what are the primitives of lexical semantic interpretation, and how they are expressed in the grammatical and morphological representation of a lexical item. The answer involves a characterization of roots as theoretical objects, followed by an argument to the effect that it is not roots, but larger structures of variable size which relate to lexical concepts. An in-depth discussion of nouns leads to the claim that the conceptual content of a lexical item does not reflect its grammatical structure, because a concept is not the meaning of a linguistically defined unit, but a language-external cognitive content, globally associated with the lexical word as a whole.
In diachronic lexicology, word formation (henceforth: WF) is generally treated as one of three strictly distinct types of processes of lexical innovation, the two others being meaning change (MCh) and borrowing. A unitary theoretical framework based on Construction Grammer and integrating different types of lexical innovation will show that, from an onomasiological point of view, WF and MCh in particular are only two ways of switching from a construction 1 expressing a concept C1 to a formally and semantically related construction 2 expressing a concept C2. From this perspective it is not surprising to find multiple combination, intertwining and interaction between MCh and WF: WF and MCh as subsequent steps in chains of diachronic change; WF and MCh as designative alternatives in lexical innovation; WF and MCh as components of the phenomena of lexical ellipsis and folk-etymology; MCh governing the diachronic evolution of already existing WF devices; MCh creating new WF devices.
Morphologically and syntactically conditioned alternations indirectly code their conditioner. The distinction between semantic information that is coded by a property of an expression and semantic information that conditions and constrains a property of an expression brings rigor into linguistic description and makes us understand an important mechanism of interpretation and reanalysis of linguistic structure. The paper provides a theoretical basis for describing indirect coding both on the paradigmatic and on the syntagmatic axis, but then focuses on syntagmatically mediated coding.
In this paper we discuss two cases of seemingly polysemous derivational affixes: the Ewe suffix -ví, originally a noun meaning “child”, which has acquired a number of different semantic values in word formation, as e.g. “inexperienced” (núfíáláví “inexperienced teacher”) or “person who adheres to the typical behaviour of a group” (amredzóví “someone who behaves like a foreigner”, and the Mandarin Chinese suffix -bā, originally a lexical morph meaning “bar”, which also helps to form locative nouns conveying a broad range of meanings, as e.g. in yóuxìbā (game-bā) “amusement arcade” or in yǎnbā (eye-bā), a kind of optometry clinic. We shall show that apparent polysemy may be a consequence of generalization, rather than of specialization in meaning, and that the mechanisms involved in the evolution of derivational affixes are mostly analogous to those of grammaticalisation.
The paper investigates the semantics of non-suffixed adjectival adjective-noun (AN) compounds in Slavic languages (especially in Slovak and Czech) within a lexeme-based approach in morphology (Aronoff 1976, 1994; Fradin 2003). Two types of AN compounds exist in Slavic languages: a suffixed one, where the suffix has scope over the whole complex base (such as vysok-o-škol-ský), and a “suffixless” one, where the lexical category is marked by means of an inflectional ending (such as modr-o-ok-ý corresponding to the well-known type of compounds like blue-eyed). The article shows, on the basis of the data drawn from large corpora, that there is a basic semantic difference between the two types and it attempts at formulating the principal constraint involved in the formation of the suffixless compounds. The paper puts forward a Principle of Integrated Meronymy as the basic semantic criterion that regulates the possibility of formation of this type of adjectival compounds.
Emerging evidence suggests that integrating the constituents of compound words involves semantic composition and that this meaning construction process draws on relation information linking the constituents. Research with novel compounds (for which semantic composition is obligatory) has found that relation structures compete for selection during semantic composition and that increased competition results in increased processing difficulty. The current project investigates whether relation competition occurs in the processing of established transparent and opaque English compounds. The results indicate that more relation competition is associated with more difficult processing of compound words, even those that are semantically opaque. This indicates that a relation-based semantic composition process is initiated during the processing of established compounds, even for semantically opaque compounds where the final interpretation cannot be relational. Understanding the semantic composition process is critically important in creating a complete theory of compound processing.
The abessive suffix of the Mordvin languages (Erzya and Moksha) expresses the meaning “without”. Diachronically, the suffix presents an example of a process of grammaticalization in which a derivational ending has developed into a case suffix. However, the process has not been completed, as the derivational meaning has also been maintained in both Erzya and Moksha. Consequently, in contemporary Mordvin languages the abessive form has a twofold character: it can be used both adverbially (as a case form) and adnominally (as an adjective derivation). This article investigates both the derivational and inflectional aspects of the abessive in the Mordvin languages and compares its uses with some other cases.
Most word-formation theories assume that derivatives in general tend towards canonicity, i.e. one-to-one correspondence between form and meaning and thus full morphosemantic transparency. However, form-meaning mismatches actually are widespread, both in terms of languages and word-formation rules. These mainly fall into two types: over-marking and under-marking. In this paper we propose a classification of these deviations, distinguishing between a derivational and a lexical level. We illustrate this classification with examples from French and other languages (English, Italian and Dutch). We sketch a unified analysis of these deviations within a word-based framework. We propose to analyse the relative importance of canonicity and discrepancies in word formation from the perspective of the interaction between the speaker and the hearer.
Nominal compounds and corresponding phrases provide a suitable test bed for inquiry into the interface and demarcations between syntax and morphology. In this article we show that the distinction between the two constructions is blurred from a theoretical perspective and investigate processing and cognitive aspects. After an examination of structural as well as semantic properties of the two types, we report on three studies (memorization, questionnaire, and reading time) that experimentally juxtapose German adjective-noun constructions of phrasal and morphological kinds. The results of these studies indicate cognitive differences between compounds and phrases. Considering alternative explanations, we argue that our data supports the view of two separated structural levels of grammar.
We report on a psycholinguistic study of semantic transparency among English compound words. We employed the P3 technique (Libben et al. 2012), which combines elements of three experimental paradigms: progressive demasking, naming, and word typing. Both the progressive demasking data and the word typing data showed a graded effect of semantic transparency associated with the number and location of semantically opaque constituents in the compound. Typing duration effects were evident at the constituent boundary, supporting observations first made by Sahel et al. (2008). We suggest that these data point to the value of letter typing durations in the analysis of morphological processing and the importance of a psychocentric perspective on lexical processing – one that emphasizes the psychological nature of morphological structures.
This paper examines the interaction between form and meaning in creole word-formation, drawing on evidence from Kriyol, a Portuguese-based creole spoken in Guiné-Bissau. The goal will be to illustrate how full reduplication interacts in complex ways with other morphological operations such as conversion, derivation and inflection. A formal morphological analysis of the layering between form and meaning will be sketched, within Construction Morphology, which treats full reduplication as a genuine lexeme-formation process. Within this word-based theory, the interaction between morphological operations is captured through the unification of construction schemas.
Compounds such as actress-cum-model, pub-cum-hotel or military-cum-religious represent mainly an alternative construction to coordinative compounds of the singer-songwriter type. Although most of them are double-headed, some can show semantic left-headedness. We have detected six main semantic groups out of a corpus of 300 types of -cum- nominal compounds. They correspond, in descending order of frequency, to people’s professional or leisure activities; buildings, rooms, and other places; literary and audio-visual genres; pieces of furniture, clothes, and other objects; abstract nouns; and toponyms, and anthroponyms. Under the heading “People”, we consider two main subheadings: team-sports players and antonymous occupations or functions.
This contribution studies the synchrony and diachrony of German plural variants. Plural doublets in the sense of overabundance must be restricted to variants with the same meaning used by one and the same speech community. But there is a preference for distinctiveness which assigns different pragmatic or semantic meanings to variants. This includes also a preference for iconic relationships between form and meaning. The assumption of a preference for biuniqueness appears to allow greater generalizations than the ideal of canonical paradigm forms.
Noun Incorporation Constructions commonly involve one noun root and one verb root, but in some languages the verbal member may be compounded with two nominal roots, which almost always have the semantics of a theme and a non-theme (a problematic fact for Baker’s 1988 theory). Moreover, if the non-theme noun specifies instrument or manner, the linear order in which the two nouns appear may vary, one noun taking scope over the other. This fact too can be a problem for Baker’s theory, as well as for any theory making use of a fixed Thematic Hierarchy. By reviewing the most salient features of Noun Incorporation Constructions in general, a theoretical apparatus is proposed which dispenses with the notion of head movement in favor of XP movement, yet preserves the idea of a Thematic Hierarchy by means of a phase-based model of word formation.
The central claim of this paper is that, besides phonological optimization, a major function of shortening words by different techniques such as clipping or acronymy is a controlled reduction of morphosemantic transparency in lexical domains, where partly or entirely opaque words are preferred to completely transparent ones. E.g., the functions of uniquely identifying and individualizing a referent by a proper name are better fulfilled by opaque labeling than by a transparent descriptive nomination. In fact, a study of shortening techniques used for official and commercial proper names contrasted with non-onymic words from the general lexicon and certain jargons in German, Farsi and Standard Chinese, three languages extremely different with respect to linguistic type and writing system, reveals a clear contrast between highly transparent shortenings preferred in the non-onymic lexicon versus less transparent or even opaque ones in the proper names, although the preferred shortening techniques themselves vary from language to language.
The paper concerns the structural factor of ‘root transparency’ in relation to the morphology-meaning interface in the mental lexicon of Hebrew. Semitic root structure was analyzed by comparing responses of native speakers of Hebrew on a set of written tasks. Target items included both nouns derived from transparent full triconsonantal roots (e.g. migdal “tower” from the root g-d-l) and nouns derived from more opaque defective biconsonantal roots (e.g. taxazit “forecast” from the root x-z-y). Responses to items based on full roots demonstrated that they were across-the-board more accessible and easier to process than their defective counterparts. Moreover, the difference between performance on items derived from full compared with defective roots was found to interface with phonology, semantics, and syntax, as well as with familiarity/frequency. These results indicate that root transparency plays an important and pervasive role in the mental lexicon of Hebrew.
This paper will discuss several lexical semantic differences between mimetic verbs and conventional prosaic verbs in Japanese, and show that meaning for mimetic verbs calls for a treatment distinct from that of lexical verbs. Taking into account the notion of ‘affect-imagistic dimension’ (Kita 1997) in which a mimetic base is represented in terms of its images, I will argue that an analysis that incorporates the basic premise of frame semantics and construction grammar can accommodate the seemingly idiosyncratic and unconventional lexical semantic properties of mimetic verbs. The wide range of semantic frames of mimetic verbs is thus fed by the numerous images of the mimetic base and the specific constructions in which they appear.
The term ‘deponent’ traditionally refers to a class of verbs in Latin or Greek which is characterized by a mismatch between their morphological form and their syntactic and semantic context. Deponent verbs appear in passive form but in active syntax/semantics. In this paper, I argue that this mismatch between form and function is not as uncommon as one might expect. In particular, I show that we find the same kind of mismatch with unaccusative verbs in many languages amongst which are English or the Romance languages. The argumentation will be based on a detailed investigation of how deponent and unaccusative verbs behave in all modules of the grammar. Based on the findings of this discussion, I propose a unified morphosyntactic analysis for deponents and unaccusatives. The analysis elegantly captures the particular behavior of these verbs by invoking two principles: Identity Avoidance and Lexical Override.
The present paper discusses deponent verbs in Modern Greek. Deponents are claimed to be idiosyncratic, non-canonical verbs representing a mismatch between morphology and syntax/semantics, as they are intransitive verbs exhibiting a morphological form, non-active, without there being an active transitive counterpart. Since non-active morphology is standardly associated with de-transitivization, deponents are taken to be exceptional in this respect. We compiled a deponent-verb corpus of Modern Greek and examined its morphological, syntactic and semantic structure. The results of this study revealed that most deponent predicates are actually reflexives/reciprocals, anticausatives and passives; thus, they instantiate verbal alternations which typically surface with non-active morphological marking. For this reason, we conclude that the morphology of deponent verbs in Modern Greek has, contrary to their traditional definition, a canonical function.