This course provided us with a comprehensive and practical introduction to the research methods used in the LUCL Labs, including the chance to consolidate this knowledge by carrying out an independent research project.
We learned how psycholinguistic theory relates to experimental design and gained practical experience of data collection and interpretation by carrying out a replication study.
From Face to Fingers:
Where Do NGT Signers Look?
Dailenn Thieme
Methods in Experimental Linguistics
Drs. Tingting Zheng
November 2025
Gestures accompanying language are as old as spoken language (DawnSignPress, 2016). The first recorded instance of sign language was in the 17th century but sign languages have only started to be recognised as fully fledged languages around the 1960s:
Perhaps the most important development over the past century has been the recognition and acceptance of natural sign languages, such as American Sign Language, as fully formed human languages distinct from, yet linguistically parallel to the surrounding spoken languages in which they exist. This recognition can be largely traced to the work of William Stokoe and colleagues (Stokoe, Casterline, & Croneberg, 1965). (Higgins & Lieberman, 2016, p. 11)
In the past 60 years, American Sign Language (ASL) and a few others such as British Sign Language (BSL) have been studied extensively. One main topic of interest in ASL research is visual attention. One study investigated the visual attention of young ASL signers comparing focus on signs versus objects in the visual world that are topics of conversation (MacDonald et al., 2018). This kind of research is unique to sign languages as sign languages require a visual channel.
Gaze behaviour is a crucial part of visual attention. Research into gaze behaviour shows that there is a significant pattern of gaze fixation that gives us insight into how people process sign language (Emmorey, Thompson & Colvin, 2009). For a significant percentage of time, 61%-99%, signers focus on the face of the person signing to them instead of their hands. Although three factors would influence the gaze in a shift away from the face and towards the hands or body: signs close to the face, expensive signs in the lower body region and movement of the signer. There have also been ASL studies on gaze behaviour using non-typical signs, which investigated videos showing signing in reversal (Bosworth, Stone & Hwang, 2020) or non-linguistic movements (Bosworth, Hwang & Corina, 2022). They found that gaze behaviour is less focused on the face in these conditions than during typical signing for fluent ASL signers.
Gaze behaviour has also been studied on other sign languages such as British (BSL) and German (DGS) Sign Language. Both BSL and DGS research found that signers of those languages tend to focus more on the face (Muir & Richardson, 2005; Parr & Zeng, 2021; Sander et al., 2025). Muir & Richardson argued that the fixation on the face region in BSL enables signers to pick up small detailed movements such as facial expressions and mouth shape while peripheral vision is used to discern the hands, which contain larger, rapid movements. Sander et al., found that in DGS signers shift their attention away from the face and towards the hands when processing unfamiliar signs.
Due to historical and political reasons, Dutch Sign Language (NGT) has not been extensively researched. A little-known fact is that NGT was banned between 1880 and 1980 (Stroosnijder, 2024). NGT was officially recognised in 2021, establishing a standardised sign language in the Netherlands. Some research has been done on gaze behaviour in NGT. Research has investigated gaze competition during message preparation and the role of eyebrows in NGT by analysing gaze fixations (Manhardt et al., 2020; De Vos, Van der Kooij & Crasborn, 2009).
However, there is no research that investigates areas of interest (AOIs) in NGT and possible influences thereupon. As mentioned earlier, there are a multitude of conditions that draw visual
attention away from the face. One condition that has not been researched is the influence of morphological complexity. NGT morphology uses simultaneous morphological processes rather than purely sequential affixation (Pfau & Steinbach, 2023). This means that the simultaneous realisation of morphosyntactic features can be manipulated concurrently to express grammatical meaning. Furthermore, some signs contain non-manual features such as mouth movements or facial expressions. There is a possibility that morphological complexity influences gaze behaviour in NGT and other sign languages.
Most of the studies mentioned above use an eye-tracking method. Eye-tracking is especially effective when studying sign languages, as they rely entirely on visual input. Eye-tracking can measure fixation patterns and duration of a viewer’s gaze. The results provide precise temporal data on visual attention. In sign language research, eye-tracking can distinguish AOIs on which viewers focus and how visual attention can shift during different signs.
In this study we aim to investigate the gaze fixation patterns in NGT and whether morphological complexity of signs influences these patterns. We hypothesise that we will find similar patterns as have been found in ASL, BSL and DGS in fluent NGT signers, with over 70% of gaze distribution focused on the face. Our second hypothesis predicts that more complex morphological signs are more likely to draw gaze attention away from the face and to the hands. Previous research has shown that unfamiliar signs, reversed signing, and expensive signs draw attention to the hands. If this is the result of peripheral vision not being enough to discern all information accurately, we predict that morphologically complex signs are also more difficult to process solely with peripheral vision. We predict that our results will show that a high percentage of time, signers will look at the face of the person signing to them, but that this percentage will slightly decrease as morphological complexity increases.
This study will address a gap in research on NGT, a language that is currently severely under-investigated. We hope that more research in NGT will not only contribute to the scientific
understanding of this language but also increase awareness, as only an estimated 60,000 are able to use sign language, even though there are approximately 1.5 million Deaf people in the Netherlands (Van der Most, 2025). Furthermore, research in NGT can help provide a foundation to build a framework upon, and promote cross-linguistic insights into sign language structure and processing.
METHODS
This study will be approved by the ethics committee of Leiden University before any experiment is conducted.
PARTICIPANTS
We will collect data from 30 native NGT signers, aged between 21 and 50 years old, equally distributed among both genders. Almost all people living in the Netherlands are familiar with Dutch and English, but no exclusions will be made based on multilingualism, as long as they are fluent in NGT. All participants require normal or corrected-to-normal vision and no diagnosed intellectual disabilities. All personal information will be anonymised. Participants will sign a consent form before the experiment and it will be made clear that they can exit the experiment at any moment. Up until the data is analysed, participants can request their data to be destroyed.
MATERIALS
Video clips will show signs with three levels of morphological complexity (see table 1). The lowest level is fingerspelling. Fingerspelling is the signing of words not by a single gesture, as is most common, but by signing out each letter individually. Fingerspelling is often used for names and words that do not have a sign yet. The second level of complexity is grammatical inflection, where signs indicate aspects such as a temporal, aspectual or numerical inflection. For example, pluralisation of words (for more research on morphological reduplication, see Van Boven (2025)) or agency of verbs. See figures 1 and 2 for the difference between “ik waarschuw jou” I warn you and “jij waarschuwt mij” You warn me. The hand gesture is the same, but the directionality influences agency. The
third level of morphological complexity is simultaneous morphology, where morphological indicators are presented concurrently. For example, in NGT, different eyebrow configurations are used to indicate polar or content questions while using the hands for lexical signs (De Vos, 2009). In figure 3, a woman is denoting a question by raising her eyebrows while signing the word meeting. The target signs will be hidden within a simple sentence to mask them and the video clips will be randomised. The clips will all be made by a native NGT signer. In total there will be 60 video clips (20
clips x 3 conditions), split into two runs lasting around 10-15 minutes. Each participant will watch all clips. After each clip, the participant will be asked a brief question to confirm they understood the signs.
Table 1
Independent variables
Morphological complexity AOI
Level 1: fingerspelling Area 1: lower upper body
Level 2: inflection Area 2: upper upper body
Level 3: simultaneous morphology Area 3: face
Figure 1: “Waarschuw mij” [A woman signing warn me] https://ow.gebarencentrum.nl/gloss/-715965393/0
Figure 2: “Ik waarschuw jou” [A woman signing I will warn you] https://ow.gebarencentrum.nl/gloss/-715965393/0
Figure 3: Meeting? [A woman signing the word meeting within a question] https://ow.gebarencentrum.nl/gloss/-715965393/0
SETUP
Participants will sit in front of a screen with their head resting on the headrest. The eye-tracking method uses a small illuminator that shines infrared light into the eye of the participant. This way, the software can track the gaze direction of the eye. We will use the EyeLink 1000 eye-tracking system at Leiden University.
Participants will watch a video clip in which a woman is signing in the middle of the screen. Each clip will last around 5 seconds. The person signing will be in the middle of the screen, shown from the waist up so that all AOIs will be in frame but the areas are as large as possible to ensure the most accuracy.
Each trial begins with a fixation marker lasting 500ms before showing the clip. Participants are encouraged to take a break between the runs for one minute by closing their eyes while remaining in the eye-scanner. If a participant feels the need to leave the scanner during a break, they will go through a recalibration sequence before resuming the experiment.
DATA PRE-PROCESSING
The eye-tracking and the screen recording data will be extracted and annotated using annotation software, such as ELAN. The annotation scheme will include fixations, blinks, AOIs, and exclusion categories, such as looking off-screen and loss of tracking. These annotations will provide the basis for the data analysis.
The gaze replay videos and the stimuli video clips will be uploaded to the Open Science Framework. The recordings of participants’ eyes will not be made public to ensure privacy.
ANALYSIS
From the raw data we will discern three AOIs, the lower part of the upper body, the upper part of the upper body and the face. The percentage of gaze fixation will be calculated across these three parts. An eye-tracking gaze plot will visualise the spatial distribution and duration of participants’ fixations.
Data will be excluded from analysis if the participant did not respond correctly to the questions and misinterpreted the signs, if there is a high level of missing gaze data (>50%) or if the participant looked outside the AOIs or the screen during the entire duration of the target sign.
We will use a two-way repeated measures ANOVA, with AOI and morphological complexity as the independent variables (both containing three levels) and the gaze duration as the dependent variable. To follow up, we will need to run a pairwise comparisons test (with Bonferroni correction) to explore the interaction effect.
PREDICTED RESULTS
We predict that more than 70% of gaze fixations will be located on the face AOI for all three levels of morphological complexity. The statistical test will show that this is a significant amount of time. There will be a slight decrease in the face AOI and an increase in the upper body AOI for the second level of morphological complexity, inflection, but we predict that this will be a trend among participants but not significant according to statistical tests. The third level, simultaneous morphology, will not show a significant or a noteworthy trend regarding decrease in fixation on the face AOI, as there is more subtle but highly important information conveyed through facial expressions. We predict a small AOI × complexity interaction effect due to the slight shift to the body AOI for inflected signs, but not strong enough to be significant.
DISCUSSION
Research into NGT is behind compared to research into other sign languages such as ASL and BSL. An important aspect of sign language comprehension is visual attention. Eye-tracking can give us insight into the gaze behaviour of signers. We will investigate the gaze behaviour of native NGT signers across different morphologically complex signs. Research into NGT will contribute to the gap in the research of this underappreciated language. If the currently predicted results are found, this would imply that facial cues and mouth movements carry great importance in NGT, similar to other sign languages, and that morphological complexity is an area of research that should be investigated further.
We argue that confirming similarities between NGT and other, well researched sign languages can help us build towards a framework for NGT and promote cross-linguistic insights into sign language structure and processing. We will investigate the effects of morphological complexity on the gaze distribution across AOIs. In ASL, BSL and DGS, researchers have found that most of the time, people are focused on the face instead of the hands. This research also contributes to broader debates on how language modality shapes cognitive processing, particularly in relation to morphological encoding and visual attention.
Conversely, it is also possible we do not find significant results. If this is the case, we must emphasise that it is entirely possible for NGT to work differently to the other investigated sign languages. This could be due to NGT having more complex signs, or less weight is put on facial expressions in NGT. In any case, more research into NGT is warranted.
REFERENCES
Bosworth, R. G., Hwang, S. O., & Corina, D. P. (2022). Visual attention for linguistic and non-linguistic body actions in non-signing and native signing children. Frontiers in Psychology, 13, 951057
Bosworth, R., Stone, A., & Hwang, S. O. (2020). Effects of Video Reversal on Gaze Patterns during Signed Narrative Comprehension. The Journal of Deaf Studies and Deaf Education, 25(3), 283–297. https://doi.org/10.1093/deafed/enaa007
De Vos, C., Van der Kooij, E., & Crasborn, O. (2009). Mixed signals: Combining linguistic and affective functions of eyebrows in questions in Sign Language of the Netherlands. Language and Speech, 52(2-3), 315-339. https://journals.sagepub.com/doi/pdf/10.1177/0023830909103177
ELAN (Version 7.0) [Computer software]. (2025). Nijmegen: Max Planck Institute for Psycholinguistics, The Language Archive. Retrieved from https://archive.mpi.nl/tla/elan
Emmorey, K., Thompson, R., & Colvin, R. (2009). Eye gaze during comprehension of American Sign Language by native and beginning signers. The Journal of Deaf Studies and Deaf Education, 14(2), 237–243. https://doi.org/10.1093/deafed/enn037
Higgins, M., & Lieberman, A. M. (2016). Deaf students as a linguistic and cultural minority: shifting perspectives and implications for teaching and learning. Journal of Education, 196(1), 9-18.
DawnSignPress (2016). History of American Sign Language. https://www.dawnsign.com/news-detail/history-of-american-sign-language
MacDonald, K., LaMarr, T., Corina, D., Marchman, V. A., & Fernald, A. (2018). Real-time lexical comprehension in young children learning American Sign Language. Developmental Science, 21(6), e12672. https://doi.org/10.1111/desc.12672
Manhardt, F., Özyürek, A., Sümer, B., Mulder, K., Karadöller, D. Z., & Brouwer, S. (2020). Iconicity in Spatial Language Guides Visual Attention: A Comparison Between Signers’ and Speakers’ Eye Gaze During Message Preparation. Journal of Experimental Psychology: Learning, Memory, and Cognition, 46(9), 1735–1753. https://doi.org/10.1037/xlm0000843
Muir, L. J., & Richardson, I. E. G. (2005). Perception of Sign Language and Its Application to Visual Communications for Deaf People. The Journal of Deaf Studies and Deaf Education, 10(4), 390–401. https://doi.org/10.1093/deafed/eni037
Parr, N., & Zeng, B. (2021). British Sign Language (BSL) user’s gaze patterns between hands and face during online communication. Companion Publication of the 13th ACM Web Science Conference 2021 (pp. 10-14). https://doi.org/10.1145/3462741.3466645
Pfau, R., & Steinbach, M. (2023). Morphology in Sign Languages. The Wiley Blackwell companion to morphology, 1445-1482
Sander, J., Meister, N. K., Finkbeiner, T. A., Rowland, C. F., Steinbach, M., Friederici, A., Zaccarella, E.,
& Trettenbrein, P. (2025). Deaf signers adapt their eye gaze behaviour when comprehending an unknown sign language. PsyArXiv. doi:10.31234/osf.io/nprju_v1
Stokoe, W. C., Casterline, D. C., & Croneberg, C. G. (1965). A dictionary of American Sign Language on linguistic principles. Washington, DC: Gallaudet College Press.
Stroosnijder, Y. (2024). De geschiedenis van de Nederlandse Gebarentaal: het ontstaan, het verbod
en de erkenning. Doof. https://www.doof.nl/samenleving-maatschappij/de-geschiedenis-
van-de-nederlandse-gebarentaal/
Van Boven, C. (2025). Morphological reduplication in Sign Language of the Netherlands: A typological and theoretical perspective. Sign Language & Linguistics, 28(1), 149–162. https://doi.org/10.1075/sll.00093.van
Van der Most, L. (2025, March 17). Introduction to Dutch: The Netherlands’ official language. Expatica Netherlands. https://www.expatica.com/nl/education/language-learning/dutch-language-123501/