Meetings

A heritage speaker by any other name: Variation in speech fluency

Onur Özsoy from HU will give a talk on the differences between individuals in their mono- and bilingual production of language. We'll go out afterwards as always!

Where are we now and how do we move forward?  A case study of the reproducibility of articles published in the Journal of Memory and Language under the open data policy

Anna Laurinavichyute from Uni Potsdam will start off this academic year. We'll go out afterwards as per tradition :)

In 2019 the Journal of Memory and Language instituted an open data and code policy; this policy requires that, as a rule, code and data be released at the latest upon publication. How effective is this policy? We compared 59 papers published before, and 59 papers published after, the policy took effect. We looked at the data sharing rates and whether papers published under the open data policy were reproducible. We'll discuss the results of the evaluation and suggest how to improve reproducibility of published research.

Predicting Compound Words for Imaginary Objects

We're having our guest speaker Laura Isabel Raveling (HU Berlin) tell us all about her recent work. We'll go out afterwards as per tradition :)

In the present work, we investigate which (novel) words speakers use to denote (novel) objects. This may include choosing a non-existing word to label novel stimuli, with compounding being the most productive mechanism involved in this process. It has been hypothesized that distributional semantic models entail cognitively plausible label candidates for concepts. As a consequence, it should be possible to employ these models to create computational simulations of human sender-receiver behavior for compound words. In order to explore that assumption, we are comparing statistically predicted semantic image descriptions with conversational label choices. We set up a generative adversarial neural network and a linear regression model to produce a semantic label embedding for a set of visual objects. Both models should be able to produce a word – including novel compound words – that closely mimics human language production. The current results show that the statistical models reliably predict semantic information present in the visual data.

Assessing the replication landscape in experimental linguistics

The speaker will be Kristina Kobrock from Uni Osnabrück. She'll be talking about her new paper with Timo Roettger in Glossa: Psycholinguistics that you're all invited to read and discuss. We'll also be hanging out afterwards!

Replications are an integral part of cumulative experimental science. Yet many scientific disciplines do not replicate much because novel confirmatory findings are valued over direct replications. To provide a systematic assessment of the replication landscape in experimental linguistics, the present study estimated replication rates for over 50,000 articles across 98 journals. We used automatic string matching using the Web of Science combined with in-depth manual inspections of 274 papers. The median rate of mentioning the search string “replicat*” was as low as 1.7%. Subsequent manual analyses of articles containing the search string revealed that only 4% of these contained a direct replication, i.e., a study that aims to arrive at the same scientific conclusions as an initial study by using exactly the same methodology. Less than half of these direct replications were performed by independent researchers. Thus our data suggest that only 1 in 1250 experimental linguistic articles contains an independent direct replication. We conclude that, similar to neighboring disciplines, experimental linguistics replicates very little, a state of affairs that should be reflected upon.

Czech number agreement attraction: Modifying attractors with relative clauses

The speaker will be Radim Lacina from Uni Potsdam and Uni Osnabrück. Grab some food and a beer with us afterward!

Agreement attraction effects in comprehending sentences such as *The key to the cabinets were rusty have long been documented across many languages. Researchers have put forward cue-based retrieval models to explain them. However, recent studies on Czech show that in the language, agreement attraction is either not present at all or tiny in size (Lacina & Chromý, 2022; Chromý, Lacina, & Dotlačil, under review). In the current study, we test whether the effects in the language could be strengthened by increasing the degree of feature-match between the verb and the attractor. We attempt to do this by modifying the attractor with a relative clause, thus giving it the subject feature. We ran a web-based moving window word-by-word self-paced reading study with Czech comprehenders (N = 280) and analysed the post-verbal region RTs with Bayesian models, where we found evidence of an attraction effect. We compared our estimates to those of Exp 4 in Chromý et al. (in revision). Contrary to our hypothesis, the data showed a very similar effect to that found previously. In future work, we aim to test whether simply increasing the distance between the attractor and the verb produces similar effects

Diversity in child language acquisition research

The speaker will be Rowena Garcia from Uni Potsdam and MPI for Psycholinguistics. Grab some food and a beer with us afterward!

Her presentation will be based on these two articles:

• Kidd, E., & Garcia, R. (2022). How diverse is child language acquisition research?. First Language.

• Kidd, E., & Garcia, R. (2022). Where to from here? Increasing language coverage while building a more diverse discipline. First Language.

Mouse tracking for on-line processing: a focus on analysis

The speaker will be Carlotta Zona from Uni Potsdam. Grab some food and a beer with us afterward!

Mouse tracking (MT) is a paradigm using computer-mouse movements as a means of measuring the relative motor attraction, or cognitive activation, of alternative responses during decision making (Freeman, 2011; Freeman et al., 2018). Studies using MT typically rely on aggregated summary measures (e.g., MAD, AUC), which might decrease statistical power (Barr, 2008) and introduce statistical artifacts (Wulff, 2019). In this talk, I give a brief overview of the statistical approaches to modeling on-line changes in MT measures for psycholinguistics research, with a focus on growth-curve analysis (Mirman, 2014) and cluster-based permutation (Barr, 2008). I conclude with a brief presentation of example results from my own research.

When machines become journalists: a linguistic investigation into metaphors in algorithmically generated football match reports.

Mariya Hristova will present her master's thesis work. Join us for a beer afterward!

Paper: https://www.researchgate.net/publication/364715346_Wenn_Maschinen_Journalisten_sind_Eine_linguistische_Untersuchung_von_Metaphern_in_algorithmisch_generierten_Fussballspielberichten 

Beyond means: distributional analyses 

Laura Ciaccio will present a new approach to analyzing data focusing not only on means but also on other distributional properties of the data. Join us for a beer afterwards!

Paper: https://link.springer.com/article/10.3758/s13423-022-02109-w

Replicating a Visual Word eye-tracking study: Probing effects in Özge et al. (2019, 2021) in German-Turkish heritage speakers 

Onur Özsoy will introduce his latest eye-tracking project and is happy to receive some feedback. We will go for beers after the session!

HSP 2022 satellite event

https://hsp2022.ucsc.edu/

We will host a satellite viewing event on Day 2 of the 35th Annual Conference of Human Sentence Processing (formerly CUNY). This year's conference is hosted by UC Santa Cruz and is fully virtual.

TMS as a tool for investigating linguistic and pragmatic processing (hybrid)

https://hu-berlin.zoom.us/j/69605157852

Isabella Boux will introduce the transcranial magnetic stimulation method (TMS) and will illustrate its application by presenting her own study on the comprehension of indirect speech acts.

Paper discussion: Bidirectional SPR  (Zoom only)

https://hu-berlin.zoom.us/j/69605157852

Due to the Omicron situation, our first meeting in 2022 will be hosted on Zoom.

Dario Paape will introduce his new preprint exploring the method of bidirectional self-paced reading and our fellow Prime Onur Öszoy will lead the discussion.

Paper: https://psyarxiv.com/d7pvz/

Kick-off Meeting of The Friday Primes! (hybrid)

https://hu-berlin.zoom.us/j/69595874844

This will be our first meeting where we'll get to know each other and where you'll learn everything about this project. Everyone is welcome!

Our co-organiser, Daniela, will also present her work on life-time effects in comprehension.