Discoveries about Languages

CoEDL PhD student Tina Gregor uncovered the intriguing phenomenon of gender suppletion in the language Yelmek [1]. Suppletion occurs when words are not formed in the predictable way. For example, when conjugating verbs in the past tense in English, we normally add -ed to verbs — count becomes counted — but this is not the case when go becomes went. Since forming new sentences depends on using regular rules, there are usually few suppletive sets per language. Tina discovered that Yelmek suppletes for the gender of objects, something no other language is known to do. The Yelmek words poyopoa (‘he held her’) and peŋepea (‘he held him’) show this divergence, with the stems oyopo and eŋepe differing depending on the gender of the object being held. Yelmek has dozens of pairs like this.

Following CoEDL Research Fellow Don Daniels’ fieldwork on the Trans New Guinea language Soq, Daniels and PI Greville Corbett established a new phenomenon termed repartitioning [2]. Verbs in Soq have a four-way tense system, which divides time reference into the far past, the yesterday past, the ‘today until now’ present or the future. An exception to this system is the unique verb s- (‘stay’). It uses the same forms but divides time differently: the ‘yesterday past’ form means ‘yesterday until now’ and the ‘today until now’ form means ‘now’. Right meaning and wrong form — as with irregular verbs like go becoming went — is a familiar divergence but the irregularity of repartitioning — right form, wrong meaning — has not been observed before.

Engagement, in linguistics, is the use of inflections on verbs to indicate whether the speaker thinks their conversation partner is paying attention to an action they are talking about. French linguist Jon Landaburu first reported this phenomenon in the Colombian language Andoke. A major survey of engagement by Director Nick Evans and SCOPIC team members Henrik Bergqvist and Lila San Roque gathered evidence from many languages to set up a logical framework for exploring the phenomenon [3]. Since then, scientific interest in engagement has escalated. A special journal issue of Open Linguistics on ‘Explorations in Engagement’ was published in 2019, including an important article by CoEDL Research Fellow Bruno Olsson exploring new techniques for understanding engagement based on video-recorded interaction [4].

Australian languages are famous for their free word order and offer a special challenge to understanding how speakers plan their sentences. Research by CI Evan Kidd, CI Rachel Nordlinger and Research Fellow Gabriela Garrido Rodriguez — which began in CoEDL before receiving ARC Discovery funding — used mobile eye-trackers to investigate this among speakers of the Australian language Murrinhpatha. Their findings — published in the field’s premier journal, Language — indicated that Murrinhpatha speakers begin planning their sentences much earlier than English speakers, probably because they need a holistic overview of the situation before uttering their first word[5]

Various studies from CoEDL members have investigated how children learn complex language structures; for instance, Research Fellow William Forshaw’s research on polysynthetic verbs in Murrinhpatha [6, 7], or Research Fellow Hannah Sarvasy’s work on the paragraph-long sentences termed clause chains in the Papuan language Nungon [8]. Further research by Affiliate Barbara Kelly and collaborators [9] and by CI Evan Kidd and Rowena Garcia [10] shows that children learning these languages appear to master complex structures on par with, or even ahead of, English-learning children mastering equivalent structures. This work pushes the field of developmental psycholinguistics to a wider range of languages and suggests that complexity or rarity may not simply signify difficulty with acquisition.

Any theory of language dynamics must account for the individual variation between language as a shared social system and language as an individual skill or activity. CoEDL members investigated this question in different ways. CI Evan Kidd, Research Fellow Seamus Donnelly and PI Morten Christiansen assessed individual differences in learning and processing [11]; Evan Kidd and Joanne Arciuli also showed statistically that children’s very early abilities to learn individual words strongly predicted their later acquisition of syntax [12]. Research Fellow Danielle Barth and other SCOPIC project colleagues used ‘parallax corpora’ to investigate the relative contribution of a speaker’s language to how they frame utterances against a background of individual variation [13].

Multilingualism proved a fruitful area for CoEDL research. CI Alan Rumsey surveyed the traditional multilingual nature of Indigenous Australian societies [14], while Research Fellows Ruth Singer and Jill Vaughan reviewed the varied forms multilingualism takes worldwide in a special issue of Language & Communication on Indigenous Multilingualisms [15]. PhD student Catalina Torres’s thesis examined intonation and bilingual speech processes in Lifou, New Caledonia [16]. The highly-cited study by Jan-Willem van Leussen and CI Paola Escudero presents a revised model for understanding how second languages are acquired in relationship to a speaker’s first language [17]. In the context of language contact and change, CI Felicity Meakins, CI Caroline Jones and Cassandra Algy showed the impacts of a community shifting from one language to another on the way bilingual speakers in that community thought about space [18].

Research by CI Catherine Travis and PI Rena Torres Cacoullos overturned a widespread dogma that, over time, the two languages spoken in bilingual communities grow more similar as speakers streamline their production mechanisms into a single cognitive workflow. Bilingualism in the Community: Code-switching and Grammars in Contact puts forward a study of English-Spanish bilinguals in New Mexico, the United States, whose community has been bilingual for centuries [19]. This research showed that the expected convergence has not yet occurred, suggesting speakers are able to maintain distinct systems over many generations.

CI Anne Cutler (†), PhD student Jiyoun Choi and Mirjam Broersma made an important contribution to understanding how the languages we learn restructure our minds and how long this effect lasts [20, 21]. They looked at people who had been adopted from Korea to the Netherlands in the first few months of life. Although these people had not learned Korean even in childhood, the study showed that, as adults, they could still produce certain Korean sounds they had been exposed to in infancy. This suggests that bilingual knowledge acquired even at six months persists, sometimes subconsciously, into adulthood.

In a study comparing Dutch-English bilinguals and English monolinguals, Research Fellow Mark Ellison and Affiliate Luisa Miceli discovered an unconscious process they label Doppel avoidance [22]. Doppel avoidance helps bilingual speakers keep their languages apart by avoiding words that sound the same across the two languages. For example, Dutch-English bilinguals are more likely than English monolinguals to select picture over photo, since photo has a Dutch ‘doppel’ (fotograaf) but picture does not. They hypothesise that this mechanism makes it easier to monitor which language is being spoken.

Research by Director Nick Evans proposed a resolution to the debate between monogenetic (single emergence) and polygenetic (multiple emergences) perspectives on the origins of human language [23]. He proposed ‘primal multilingualism’ between half-developed languages, each with complementary innovations, lay at the heart of how language evolved. The plausibility of this theoretical model was amplified in work by CI Felicity Meakins and her collaborators, who used techniques from population biology to explore how new languages can arise by merging elements drawn from two ‘parent languages’ into elaborate systems [24]. This finding contradicts the long-held belief that the simplest elements from contributing languages are selected for as a new language emerges.

Foundational work by Research Fellow Ron Planer and CI Kim Sterelny put forward a new, gradualist account of the emergence of language as the product of over a million years of developing communications technologies and cognitive capacities. Their work culminated in the book From Sign to Symbol, which considers the insights that archaeological and biological evidence can give into linguistic history; the relationship between cognitive evolution and language use; comparisons between the communication practices of humans and our closest great-ape relatives; social and economic influences on language development; the shift between gestural to vocal modes of communication, and more [25]. This work also brought the interdisciplinary evolutionary problems of linguistic diversity to the attention of biologists and philosophers through special journal issues of Biology & Philosophy [26] and Biological Theory [27].

PI Morten Christiansen and Nick Chater showed how theories of language evolution can integrate material from observable experimental effects in their influential book Creating Language [28]. This research reveals how the iterated effects of processing and learning constraints can nudge, over hundreds of generations, the evolution of language onto particular paths, without invoking any hardwired ‘universal grammar’. The conceptual framework for integrating evolutionary theory with the systematic study of linguistic diversity is set out by Director Nick Evans, who shows how culture, demography and other factors can push linguistic evolution in different directions [29].