Periodic Reporting for period 4 - VARIKIN (Cultural Evolution of Kinship Diversity: Variation in Language, Cognition, and Social Norms Regarding Family)
Reporting period: 2020-01-01 to 2021-06-30
The VARIKIN project took a multi-disciplinary perspective to investigate the question “why do human societies differ in who they class as family?” Kinship, or who we consider to be our family, is a critical component of human social life. Every human society has rules governing our behaviour towards kin, such as who we can marry or who can inherit from whom. For decades anthropologists have documented but not fully explained the diversity in one aspect of kinship practice, kinship terminology—the language we use to refer to and classify family members. For example, some languages use different words for mother and mother’s sister, while other languages refer to these two people with the same word. The social world is carved up at different joints by denoting relatives with linguistic labels in diverse ways, but there is intriguing global consistency in the patterns. Using cutting-edge computational techniques, large linguistic and anthropological datasets, anthropological fieldwork and methodologies from evolutionary biology and cognitive science, we aimed for a comprehensive understanding of diversity in kinship terminology at the levels of child learning, everyday language use, and long-term cultural evolution.
One overarching aim was to create an open-access and comprehensive resource of kinship terms across the world’s languages. We collaborated with other research teams (from the Australian National University, Brazilian research institution Museu Paraense Emilio Goeldi, and Yale University) on this ambitious goal. We have produced KinBank: A Global Database of Kinship Terminology, which contains terms for 150 possible kin relationships across more than 1100 languages. Three main sub-projects structured our further research. Development considered what effect children’s learning constraints might have on kinship patterns; Usage studied whether the everyday use of language might pressure some kinship categories to change in different ways; and Evolution used a suite of cultural evolutionary modelling tools to understand how and why kinship terms (and their systems) change over time.
One overarching aim was to create an open-access and comprehensive resource of kinship terms across the world’s languages. We collaborated with other research teams (from the Australian National University, Brazilian research institution Museu Paraense Emilio Goeldi, and Yale University) on this ambitious goal. We have produced KinBank: A Global Database of Kinship Terminology, which contains terms for 150 possible kin relationships across more than 1100 languages. Three main sub-projects structured our further research. Development considered what effect children’s learning constraints might have on kinship patterns; Usage studied whether the everyday use of language might pressure some kinship categories to change in different ways; and Evolution used a suite of cultural evolutionary modelling tools to understand how and why kinship terms (and their systems) change over time.
The project ran from 2015-2021. Different sub-projects were led by PI Fiona Jordan and team members, but work was collaborative and cross-cutting across all strands of research.
VARIKIN-Development: How do children acquire kinship knowledge?
Previously, only a small body of research existed that considered how children acquire knowledge about kin. Most research studied English-speaking children in middle-class urban contexts. We reviewed research across anthropology, linguistics and psychology, and in 2018 held the first international workshop on this topic (CAKTAM: Children's Acquisition of Kinship Knowledge: Theory and Method), in 2018 to stimulate new empirical research, particularly with non-Western communities. To this end, Mitchell and Jordan created a research field-kit for children's acquisition of kinship terms, published on the Open Science Framework.
Dr Alice Mitchell led the Development subproject and conducted 9 months fieldwork with Datooga people in northern Tanzania. The Datooga are traditionally semi-nomadic polygynous cattle-herders. By living with a Datooga family, and using a variety of methods from anthropology, linguistics, and developmental psychology Mitchell explored how Datooga children learn about their kinship system. She video-recorded 45 hours of Datooga children's spontaneous interactions and transcribed 53,000 words from recordings, and conducted interviews with 66 children. Dr Mitchell also collaborated with Joe Blythe and Jeremiah Tunmuck (Macquarie University) who were working with indigenous Australian communities to share methodologies and results. Findings show that kinship terms are complex for young children to learn and use accurately, even in communities where much time is spent with a range of relatives.
VARIKIN-Usage: How do people use kinship language?
Language change on a historical timescale arises from modifications in the language used by individual speakers in their day-to-day speech. By detecting patterns of kin term usage in large collections of spoken and written language, we gained insight into the microevolutionary pressures that can shape linguistic change over millennia. For example, does how often people speak about certain relatives affect word change? Is there variation in how often people talk about e.g. mothers versus aunts?
Dr Peter Racz led the Usage subproject and used cutting-edge computational techniques from the field of corpus linguistics to detect patterns of kinship term use across millions of words and dozens of languages. We collated and created language corpora from 34 Indo-European languages and demonstrated two main trends. Genealogically close relatives are spoken/written about more frequently than distant ones in all the languages we studied, and the more frequently a kin term is used, the slower it is to change or be replaced in long-term language evolution.
VARIKIN-Evolution: What drives the diversity of kinship systems?
Though there is variation in kinship term systems around the world, there are also very regular patterns, and this puzzle has fascinated anthropologists for decades. We used computational and phylogenetic tools from evolutionary biology that can track shared ancestry in language to investigate these questions. Do coherent systems exist? Are they products of cultural “descent with modification”? What factors drive their variability?
Dr Sam Passmore, Dr Terhi Honkola, Dr Catherine Sheard and Prof Jordan conducted a series of cultural evolutionary investigations. Our first studies showed that kinship systems persisted over long time depths in large language families—they were slow to change. Later studies examined dozens of longstanding anthropological claims that classic “kinship systems” patterns were driven by social norms such as marriage and inheritance (Sheard, Passmore). Surprisingly, we found very little support for these claims using modern methods of analysis. We then re-examined the coherence of the classic systems themselves and showed that they too had patchy support. This is important, because these classic systems are basic concepts found in any introductory anthropological textbook. Passmore further developed new methods for inferring robust kinship systems, and Honkola addressed the question of language contact with the first global analysis and systematic review of how “borrowable” kinship terms are between languages.
VARIKIN-Development: How do children acquire kinship knowledge?
Previously, only a small body of research existed that considered how children acquire knowledge about kin. Most research studied English-speaking children in middle-class urban contexts. We reviewed research across anthropology, linguistics and psychology, and in 2018 held the first international workshop on this topic (CAKTAM: Children's Acquisition of Kinship Knowledge: Theory and Method), in 2018 to stimulate new empirical research, particularly with non-Western communities. To this end, Mitchell and Jordan created a research field-kit for children's acquisition of kinship terms, published on the Open Science Framework.
Dr Alice Mitchell led the Development subproject and conducted 9 months fieldwork with Datooga people in northern Tanzania. The Datooga are traditionally semi-nomadic polygynous cattle-herders. By living with a Datooga family, and using a variety of methods from anthropology, linguistics, and developmental psychology Mitchell explored how Datooga children learn about their kinship system. She video-recorded 45 hours of Datooga children's spontaneous interactions and transcribed 53,000 words from recordings, and conducted interviews with 66 children. Dr Mitchell also collaborated with Joe Blythe and Jeremiah Tunmuck (Macquarie University) who were working with indigenous Australian communities to share methodologies and results. Findings show that kinship terms are complex for young children to learn and use accurately, even in communities where much time is spent with a range of relatives.
VARIKIN-Usage: How do people use kinship language?
Language change on a historical timescale arises from modifications in the language used by individual speakers in their day-to-day speech. By detecting patterns of kin term usage in large collections of spoken and written language, we gained insight into the microevolutionary pressures that can shape linguistic change over millennia. For example, does how often people speak about certain relatives affect word change? Is there variation in how often people talk about e.g. mothers versus aunts?
Dr Peter Racz led the Usage subproject and used cutting-edge computational techniques from the field of corpus linguistics to detect patterns of kinship term use across millions of words and dozens of languages. We collated and created language corpora from 34 Indo-European languages and demonstrated two main trends. Genealogically close relatives are spoken/written about more frequently than distant ones in all the languages we studied, and the more frequently a kin term is used, the slower it is to change or be replaced in long-term language evolution.
VARIKIN-Evolution: What drives the diversity of kinship systems?
Though there is variation in kinship term systems around the world, there are also very regular patterns, and this puzzle has fascinated anthropologists for decades. We used computational and phylogenetic tools from evolutionary biology that can track shared ancestry in language to investigate these questions. Do coherent systems exist? Are they products of cultural “descent with modification”? What factors drive their variability?
Dr Sam Passmore, Dr Terhi Honkola, Dr Catherine Sheard and Prof Jordan conducted a series of cultural evolutionary investigations. Our first studies showed that kinship systems persisted over long time depths in large language families—they were slow to change. Later studies examined dozens of longstanding anthropological claims that classic “kinship systems” patterns were driven by social norms such as marriage and inheritance (Sheard, Passmore). Surprisingly, we found very little support for these claims using modern methods of analysis. We then re-examined the coherence of the classic systems themselves and showed that they too had patchy support. This is important, because these classic systems are basic concepts found in any introductory anthropological textbook. Passmore further developed new methods for inferring robust kinship systems, and Honkola addressed the question of language contact with the first global analysis and systematic review of how “borrowable” kinship terms are between languages.
Research collaborations allowed for a more comprehensive Kinbank than originally envisaged. These collaborations, particularly with the Parabank team at the Australian National University and the attendees of our CAKTAM workshop, stimulated further research possibilities for all team members and led to new grant funding proposals e.g. Jordan’s British Academy partnership with researchers at the Museu Goeldi, Brazil; Honkola’s KONE Fellowship at Helsinki, Finland.
As at mid-2021, many empirical results from the project have been reported in the scientific literature. 2022 will see the release of the Kinbank database. Further publications to come include our work on kinship term borrowing, on identifying robust kinship systems, and our presentation of Datooga children’s kinship knowledge. Two books are in preparation: “Relatively Speaking” (Jordan) and “The Beginnings of Kinship” (Mitchell).
As at mid-2021, many empirical results from the project have been reported in the scientific literature. 2022 will see the release of the Kinbank database. Further publications to come include our work on kinship term borrowing, on identifying robust kinship systems, and our presentation of Datooga children’s kinship knowledge. Two books are in preparation: “Relatively Speaking” (Jordan) and “The Beginnings of Kinship” (Mitchell).