The Voice System of Amharic

Desalegn passed his defense with flying colors! Those of you who have not yet downloaded his thesis can do so at Munin. I warmly recommend it for anyone interested in the syntax of argument structure, the lexical semantics of verbs, the interplay between functional structure and roots or between lexical semantics and context, the splendours of Amharic verbal morphology, causatives, passives, middles, other Voices, or any of several related topics. At 380 pages it has something to offer everybody.


Posted in Uncategorized | Leave a comment

Desalegn Workneh’s doctoral defense

We have a new doctoral defense coming up! Desalegn Workneh will defend his PhD dissertation on The Voice System of Amharic next week.

The event will start with a 45-minute trial lecture at 9AM on Thursday, September 3rd on “The interaction between viewpoint aspect and tense in Amharic morphosyntax”. All are invited to stream it live here. (The link has also been sent around by e-mail by Kari Guldahl.)

On Friday, September 4th, the defense proper will start again at 9AM. You may follow it here. (You will also find this link in that same e-mail.)

The defense opens with Desalegn summarising his thesis, followed by first opponent Dr. Mengistu Amberber who will be joining us from the University of New South Wales in Sydney (where it will already be 5PM, which is why we are starting earlier than usual) and then second opponent Prof. Michelle Sheehan, who will be joining us from Anglia Ruskin University in Cambridge, England (where it will still only be 8AM at the start of the defense).

Desalegn himself will be beaming in electronically from Addis Ababa University (where it will be 10AM at the start of each of the two events).

Don’t miss this exciting and historical event!

Posted in Uncategorized | Leave a comment

Aurora Center 2020-23

Director: Marit Westergaard

Principal investigators: Yulia Rodina & Fatih Bayram (Theme 1), Merete Anderssen & Natalia Mitrofanova (Theme 2), Øystein Vangsnes & Terje Lohndal (Theme 3), Jason Rothman & Evelina Leivada (Theme 4)


Humans are unique among animals in that we have language, a complex system enabling com­muni­cation about any topic, be it past, present or future. In fact, humans are not limited to one language, but can acquire several under the right conditions. Nevertheless, bi- and multilingualism (henceforth referred to as multilingualism, unless further specified) is not an either-or pheno­menon, as multi­lingual minds may (and typically do) undergo numerous changes across the lifespan, as a result of linguistic and non-linguistic factors. This means that multilingual minds comprise dyna­mic linguistic systems, as co-existing languages affect each other in a multi­tude of ways, both in the acquisition process and beyond.

The AcqVA Aurora Center will conduct ecolo­gi­cally valid research, re­flect­ing today’s globa­lized world, where learning mul­tiple lan­guages at various points in the lifetime has become increasingly common. Our research will focus on a range of multi­lingual speaker groups and thus feed into current chal­lenges related to migration, education, and health, addressing important and yet unanswered questions for science and society. AcqVA Aurora will combine solid empi­ri­cal work with advanced theoretical (and statistical) model­ing in three domains: A) Acquisition: how multilingual minds develop in children and adults, B) Variation: how and why languages may differ considerably across indivi­duals and groups in space and time, and C) Attrition: how and why language erosion may occur over the course of the lifespan. The three domains will be studied within four cross-cutting themes, focusing on inter­related issues of multi­lingualism: 1) how linguistic and non-linguistic experiential factors shape linguistic and cognitive outcomes, 2) how multiple languages in the same mind influence each other, 3) how clo­sely related varieties co-existing in the same mind are processed, and 4) how representing and juggling multiple languages manifest and result in adaptations at the neurological and domain-general cognitive levels.

Both senior and junior scholars will be repre­sented among the PIs, which will ensure future recruitment and fulfil the intention of the Aurora call, preparing team members for further grant proposals, e.g. a Center of Excellence (CoE) from the Research Council of Norway (RCN) and European Research Council grants (ERCs). AcqVA Aurora will hire four postdoctoral fellows and four inter­national Adjunct Professors (20%), one for each theme, plus a lab manager. Further­more, we will fill a gap in our current theoretical and methodological focus by hiring one full pro­fessor with a strong track record in neuro­cognition of multi­lingualism and qualifications in neuroimaging metho­dologies (e.g. MRI), com­plementing the beha­vioral and psycho­linguistic methods already in use in our labs. This means that AcqVA Aurora will add ten outstanding scholars to the current team, making UiT a world-leading center for research on the linguistic and cognitive sciences of multi­lingualism.

Posted in Uncategorized | Leave a comment

Aurora Center

A special Christmas message from Marit:

More terrific news today! This is (almost) incredible – but it just became official:

Our proposal to become an Aurora Center has been granted!!!

This is a 4-year / 30 million NOK grant from UiT that will enable us to hire many new people and do many cool things (see below for a summary of the center proposal).

I am really happy and excited about this. This will be a true game-changer for our research.

Thanks to everyone in LAVA/AcqVA who has worked so hard for many years to build up this amazing group. And thanks to everyone who contributed in important ways to this application.


Many happy greetings,



Posted in Uncategorized | Leave a comment

Theories of features, continued

As we’re still digesting two in-depth presentations by Omer Preminger against abstract Agree as a way to analyze anaphora and other predicate-argument relations, the discussion from the Thirty Million Theories of Features workshop continues unabated in various venues — Gillian has a new blog entry entitled Define semantics, and I have one on the question of whether plural properly contains singular. And now Thomas Graf has posted a lengthy response to my blog post on his own blog, Outdex, entitled Omnivorous number and Kiowa inverse marking: Monotonicity trumps features?

Posted in Uncategorized | Leave a comment

More than thirty million theories of features

The implicit complaint in McCawley’s book title, Thirty Million Theories of Grammar, was that having many theories of something is a less satisfactory state of affairs than having few (viable) theories of it — to paraphrase Thomas Graf’s characterization of that position, if you have more than one viable description of something, you don’t fully understand it. At the Thirty Million Theories of Features workshop, I was indeed hoping we would be left with fewer than we started from. I’m not sure we are (see Gillian’s blog posts for some summaries). But Thomas Graf articulated the opposite perspective, that it can be better to have more theories of a complex object of study — in his words, if you only have one viable description, you don’t fully understand the object you’re describing.

Now we segue into Omer Preminger week, with two additional talks and discussion with Omer Preminger (Wednesday and Friday).

Posted in Uncategorized | Leave a comment

Thirty Million Theories of Features

The Thirty Million Theories of Features workshop is upon us! The hope is that by the end of the workshop, there will be fewer theories of features than we started with — as Gillian put it in her blog post, there will be blood on the floor.

Today we start off with Thomas Graf, who will ask (and perhaps answer) the question of whether features are more trouble than they’re worth!

Posted in Uncategorized | Leave a comment


OASIS – Ontology As Structured by the Interfaces with Semantics –

is a CNRS-based interdisciplinary network project connecting CASTL to linguistic centers in Paris, Barcelona, Berlin, London, and the US to study formal semantic ontology.
The first OASIS meeting, OASIS 0, took place in Tromsø;
OASIS 1 was in Paris, and the third meeting, OASIS 2, will
take place in Nantes, France, on October 16-18, 2019.

Invited speakers:

Sudha Arunachalam (New York University)
Rose-Marie Déchaine (University of British Columbia)
Nicola Guarino (ISTC – CNR)
Angelika Kratzer (UMass Amherst)
Brent Strickland (IJN – CNRS)


The OASIS conference series aims to promote conversation and
cross-fertilization across different disciplines, using ontological
questions as shared reference points. The broad questions in the
background are these:
1. What basic ontological building blocks do we use to talk and think
about the world?
2. How do these building blocks get combined?
3. How do grammatical and cognitive phenomena motivate the answers to the
first two questions?
We welcome contributions from semanticists as well as from researchers in
domains of cognition that interface with semantics.  We would like the
OASIS conferences to help foster new perspectives and to provide a forum
around which a new research community can coalesce.
OASIS 1 (Paris, November 2018)
brought together researchers from formal semantics, natural language
syntax, philosophy, psychology/psycholinguistics, language development,
neuroscience/neurolinguistics, and computational linguistics.  Some
aspects of the exchanges at OASIS 1 are summarized on the OASIS 1 site
( and on Gillian Ramchand’s blog
( ).  The range of talks at OASIS 1
gives an indication of the kinds of topics that we welcome at OASIS
conferences.  There were, for instance, talks about flexible aspects of
linguistic meaning, about categorization in verbal vs. nonverbal
populations, about the acquisition of counterfactuality and its linguistic
expression, and about the format of syntactic structure from an embodied
cognition perspective.  The OASIS credo at
lists a variety of topics relevant to the broad questions that interest

Call for submissions:

We invite submissions of abstracts for 30-minute oral presentations (+ 10
minutes discussion) on any topic pertaining to the shared interests and
assumptions of the OASIS network, as well as for poster presentations with
lightning talks. Abstracts should be submitted via the EasyChair page
submissions is 9am GMT on Saturday June 15 2019.
Abstracts must be anonymous and should be at most 2 pages (A4 or US
Letter) in length, including examples and references, using a 12pt font
with 1 inch (2.5 cm) margins on all four sides. These limitations will be
strictly enforced. A single author can send no more than one
singly-authored and one co-authored abstract, or two co-authored
abstracts. Your submission should specify whether it is to be considered
for an oral presentation or a poster. We will not accept papers that at
the time of the conference have been published or have been accepted for
Please keep in mind that this is an interdisciplinary conference, and
write your abstract accordingly. This means that the broad goals of the
research (e.g. to understand language architecture, to understand brain
architecture) and any subgoals should be mentioned. We very much welcome
work that brings attention to data from less-familiar languages; if you
propose such an abstract, keep the dataset as streamlined as possible to
give the audience a chance to understand the issues at stake. Finally, if
you work in a formal framework, your proposal must be explained in words
as well as in your formal framework.

Contact address: oasis2-info AT .

Posted in Conferences, Uncategorized | Leave a comment

Strong representation by CASTL at GLOW in Oslo

The program for GLOW 42 (Oslo, May 7-11) is out, and CASTL is well represented!

On the main program, there’s a talk by Michal Starke together with CASTL alumnus Pavel Caha along with Karen de Clercq and Guido Vanden Wyngaerd, entitled ‘How to be positive’.

In addition, there are posters by Martin Krämer together with Chris Golston (‘Diphthongs as microfeet: Deriving a typological asymmetry’) and Serge Minor and Natalia Mitrofanova (‘A competition-based account of locative modification in Russian’). Maybe we’ll be hearing practice runs for those here soon?

Furthermore, CASTL alumna Sandhya Sundaresan and former CASTL colleague Thomas McFadden have a poster together with Hedde Zeijlstra (‘Deriving selective opacity in adjuncts’). Our colleague from long ago, Anders Holmberg, has a poster with Murdhy Alshamari (‘Topic particles, agreement and movement in an Arabic dialect’).

Congratulations to all of the above on having their abstracts accepted. Have a look at the full program to see many additional exciting talks and posters.

Posted in Uncategorized | Leave a comment

A major new project at CASTL, led by Jason Rothman

Jason Rothman has been awarded a huge grant by Tromsø Forskningsstiftelse to conduct research into bilingualism in Tromsø. The project is called HELPING, which is an acronym for ‘Heritage-Bilingual Linguistic Proficiency In the Native Grammar’, and the project has a subtitle, ‘Charting and Explaining Differences’.

The project will run from August 2019 until August 2023, and has a budget of over 25 million kroner (about 3 million euros), and is bundled together with a permanent full-time position in Tromsø for Jason.

The following text comes from the project description:

The primary objective of HeLPiNG is to answer one of the most perplexing questions in bilingualism research today: Why is HLB characterized by such variation in grammatical knowledge and language use when this is not the case for monolinguals? by addressing these equally fundamental secondary objective questions:

  • (Aim 1when and why do developing monolinguals and HSs begin to diverge for the same language?
  •  (Aim 2) at what levels (under what modalities of testing) do HSs truly differ (introducing neuro (EEG/ERP) methods to this question)?
  •  (Aim 3what is the role of the (lack of) HL literacy in explaining (some) observed HS outcomes?

There are three work packages:

  • WP1 addresses the dearth of late childhood data issue, namely that most heritage bilingual research is conducted with young adults at an end-state of acquisition as opposed to development in real time.
    • It is the first methodology to address the developmental angle of heritage grammars with a unique  method that combines cross-sections tested over a 4 year period, capturing at the end data representing 15 years of development.
  • WP2 and WP3 use psycho-/neuro- linguistic methodologies (the very first brain study of its kind).
    • These methods will reveal the depth of “difference” by looking directly at how the heritage language is processed in real time and if predictive processing is qualitative similar in HSs.


Posted in Uncategorized | Leave a comment