Chris Graham | University of California, Davis (original) (raw)
Papers by Chris Graham
Since the seminal work of Greenberg (1963), there has been a clear focus by some typologists to e... more Since the seminal work of Greenberg (1963), there has been a clear focus by some typologists to explain cross-linguistic patterns and their diachronic development, often by citing various kinds of change through language contact (Trudgill 2011). This task can be difficult given the understudied nature of many of the world’s languages, as well as the scope of the endeavors, be they focused macroscopically on a wide variety of languages or more succinctly on a particular geographical area or Sprachbund. The World Atlas of Language Structures (WALS; Dryer and Haspelmath 2011, 2014) database allows for a unique new approach to older questions of contact-induced change by virtue of loci (geographic positions) of languages included primarily for visual representation on maps. Indeed, some of these visual representations are quite striking, and some were the inspiration for the work that follows, which utilizes principles from big data analyses and geographical proximity.
The present study utilizes a novel method of iteration through the entirety of the database in an effort to bridge the gap between the macroscopic and microscopic approaches. The focus is on rare syntactic types as defined by the distribution of the orders of subject, object, and verb within the database. The properties of these subsets of languages are collected, as are the properties of their geographical correlates at various distances from each subset’s loci. This allows for comparison between the properties of a language type (e.g. non-rigid object-verb languages) and those languages which are spoken within some proximity of those in the type (i.e. with some heightened probability of being in a contact situation) over an arbitrary number of linguistic features.
The results suggest that the non-rigid OV type tends to be spoken in areas with surrounding VO speakerships relative to the rigid OV type. More generally, there is some propensity for syntactic features of a rare, disharmonic type (cf. Dryer 1992; Hawkins 2014) to reflect the features of those languages spoken in proximity – a propensity which is not shared by more common types. This is taken to support the idea that syntactic change can result from language contact. Furthermore these data are compared with existing microscopic work on language contact and contact-induced change (e.g. Emonds and Faarlund 2014; Nicolai 2003) in an effort to complement them and bridge the gap between the two approaches.
Since the seminal work of Greenberg (1963), there has been a clear focus by some typologists to e... more Since the seminal work of Greenberg (1963), there has been a clear focus by some typologists to explain cross-linguistic patterns and their diachronic development, often by citing various kinds of change through language contact (Trudgill 2011). This task can be difficult given the understudied nature of many of the world's languages, as well as the scope of the endeavors, be they focused macroscopically on a wide variety of languages or more succinctly on a particular geographical area or Sprachbund. The World Atlas of Language Structures (WALS; Dryer and Haspelmath 2011, 2014) database allows for a unique new approach to older questions of contact-induced change by virtue of loci (geographic positions) of languages included primarily for visual representation on maps. Indeed, some of these visual representations are quite striking, and some were the inspiration for the work that follows, which utilizes principles from big data analyses and geographical proximity. The present study utilizes a novel method of iteration through the entirety of the database in an effort to bridge the gap between the macroscopic and microscopic approaches. The focus is on rare syntactic types as defined by the distribution of the orders of subject, object, and verb within the database. The properties of these subsets of languages are collected, as are the properties of their geographical correlates at various distances from each subset's loci. This allows for comparison between the properties of a language type (e.g. non-rigid object-verb languages) and those languages which are spoken within some proximity of those in the type (i.e. with some heightened probability of being in a contact situation) over an arbitrary number of linguistic features. The results suggest that the non-rigid OV type tends to be spoken in areas with surrounding VO speakerships relative to the rigid OV type. More generally, there is some propensity for syntactic features of a rare, disharmonic type (cf. Dryer 1992; Hawkins 2014) to reflect the features of those languages spoken in proximity – a propensity which is not shared by more common types. This is taken to support the idea that syntactic change can result from language contact. Furthermore these data are compared with existing microscopic work on language contact and contact-induced change (e.g. Emonds and Faarlund 2014; Nicolai 2003) in an effort to complement them and bridge the gap between the two approaches.
![Research paper thumbnail of Non-rigid OV languages as languages in flux: evidence of change from OV to VO [LSA 2014 Poster]](https://attachments.academia-assets.com/32718626/thumbnails/1.jpg)
](A sample of SOVX languages (i.e. SOV ∩ OVX) was collected from datapoints of the World Atlas of L... more A sample of SOVX languages (i.e. SOV ∩ OVX) was collected from datapoints of the World Atlas of Language Structures (WALS, Dryer and Haspelmath, 2011). For each of the 39 languages in the sample, datapoints for three other features were collected: Order of Adjective and Noun (F87A, Dryer, 2011a), Order of Numeral and Noun (F89A, Dryer, 2011b), and Order of Relative Clause and Noun (F90A, Dryer, 2011c).
Language parsing is a necessary part of linguistic comprehension which has traditionally been res... more Language parsing is a necessary part of linguistic comprehension which has traditionally been researched using various psycholinguistic paradigms; a lot of work has attempted to extrapolate language-specific data into universal cognitive models of parsing. This paper argues that this kind of generalization misses an important point, namely that there is a link between language type and parsing requirements which should not be overlooked. A brief overview of information encoding by grammars is given, and two possible incremental parsing methods to extract that information are aligned with typological language data. Deeper patterns involving these parsing strategies are shown to explain the existence of typologically rare word orders in some Indo-European languages (German, Dutch, Kashmiri) as well as cross-linguistically.
Since the seminal work of Greenberg (1963), there has been a clear focus by some typologists to e... more Since the seminal work of Greenberg (1963), there has been a clear focus by some typologists to explain cross-linguistic patterns and their diachronic development, often by citing various kinds of change through language contact (Trudgill 2011). This task can be difficult given the understudied nature of many of the world’s languages, as well as the scope of the endeavors, be they focused macroscopically on a wide variety of languages or more succinctly on a particular geographical area or Sprachbund. The World Atlas of Language Structures (WALS; Dryer and Haspelmath 2011, 2014) database allows for a unique new approach to older questions of contact-induced change by virtue of loci (geographic positions) of languages included primarily for visual representation on maps. Indeed, some of these visual representations are quite striking, and some were the inspiration for the work that follows, which utilizes principles from big data analyses and geographical proximity.
The present study utilizes a novel method of iteration through the entirety of the database in an effort to bridge the gap between the macroscopic and microscopic approaches. The focus is on rare syntactic types as defined by the distribution of the orders of subject, object, and verb within the database. The properties of these subsets of languages are collected, as are the properties of their geographical correlates at various distances from each subset’s loci. This allows for comparison between the properties of a language type (e.g. non-rigid object-verb languages) and those languages which are spoken within some proximity of those in the type (i.e. with some heightened probability of being in a contact situation) over an arbitrary number of linguistic features.
The results suggest that the non-rigid OV type tends to be spoken in areas with surrounding VO speakerships relative to the rigid OV type. More generally, there is some propensity for syntactic features of a rare, disharmonic type (cf. Dryer 1992; Hawkins 2014) to reflect the features of those languages spoken in proximity – a propensity which is not shared by more common types. This is taken to support the idea that syntactic change can result from language contact. Furthermore these data are compared with existing microscopic work on language contact and contact-induced change (e.g. Emonds and Faarlund 2014; Nicolai 2003) in an effort to complement them and bridge the gap between the two approaches.
Since the seminal work of Greenberg (1963), there has been a clear focus by some typologists to e... more Since the seminal work of Greenberg (1963), there has been a clear focus by some typologists to explain cross-linguistic patterns and their diachronic development, often by citing various kinds of change through language contact (Trudgill 2011). This task can be difficult given the understudied nature of many of the world's languages, as well as the scope of the endeavors, be they focused macroscopically on a wide variety of languages or more succinctly on a particular geographical area or Sprachbund. The World Atlas of Language Structures (WALS; Dryer and Haspelmath 2011, 2014) database allows for a unique new approach to older questions of contact-induced change by virtue of loci (geographic positions) of languages included primarily for visual representation on maps. Indeed, some of these visual representations are quite striking, and some were the inspiration for the work that follows, which utilizes principles from big data analyses and geographical proximity. The present study utilizes a novel method of iteration through the entirety of the database in an effort to bridge the gap between the macroscopic and microscopic approaches. The focus is on rare syntactic types as defined by the distribution of the orders of subject, object, and verb within the database. The properties of these subsets of languages are collected, as are the properties of their geographical correlates at various distances from each subset's loci. This allows for comparison between the properties of a language type (e.g. non-rigid object-verb languages) and those languages which are spoken within some proximity of those in the type (i.e. with some heightened probability of being in a contact situation) over an arbitrary number of linguistic features. The results suggest that the non-rigid OV type tends to be spoken in areas with surrounding VO speakerships relative to the rigid OV type. More generally, there is some propensity for syntactic features of a rare, disharmonic type (cf. Dryer 1992; Hawkins 2014) to reflect the features of those languages spoken in proximity – a propensity which is not shared by more common types. This is taken to support the idea that syntactic change can result from language contact. Furthermore these data are compared with existing microscopic work on language contact and contact-induced change (e.g. Emonds and Faarlund 2014; Nicolai 2003) in an effort to complement them and bridge the gap between the two approaches.
A sample of SOVX languages (i.e. SOV ∩ OVX) was collected from datapoints of the World Atlas of L... more A sample of SOVX languages (i.e. SOV ∩ OVX) was collected from datapoints of the World Atlas of Language Structures (WALS, Dryer and Haspelmath, 2011). For each of the 39 languages in the sample, datapoints for three other features were collected: Order of Adjective and Noun (F87A, Dryer, 2011a), Order of Numeral and Noun (F89A, Dryer, 2011b), and Order of Relative Clause and Noun (F90A, Dryer, 2011c).
Language parsing is a necessary part of linguistic comprehension which has traditionally been res... more Language parsing is a necessary part of linguistic comprehension which has traditionally been researched using various psycholinguistic paradigms; a lot of work has attempted to extrapolate language-specific data into universal cognitive models of parsing. This paper argues that this kind of generalization misses an important point, namely that there is a link between language type and parsing requirements which should not be overlooked. A brief overview of information encoding by grammars is given, and two possible incremental parsing methods to extract that information are aligned with typological language data. Deeper patterns involving these parsing strategies are shown to explain the existence of typologically rare word orders in some Indo-European languages (German, Dutch, Kashmiri) as well as cross-linguistically.