Uncovering the Role of Sleep in the Acquisition of Linguistic Knowledge
Lead Research Organisation:
Royal Holloway University of London
Department Name: Psychology
Abstract
One remarkable aspect of human learning is our ability to build general knowledge from individual experiences. This general knowledge is central to virtually all cognitive functions, but is particularly important in language, as it allows us to use new words, phrases, and sentences that have not been communicated previously. For example, we understand the novel word 'untweetable' because we have general knowledge of the functions of affixes {un} and {able}. However, despite the significance of this form of knowledge for human communication, we know remarkably little about how it is acquired.
Recent research provides strong clues that sleep may play a vital role in the acquisition of general linguistic knowledge. In a previous ESRC project, we developed a laboratory analogue of language learning to track how general knowledge is built through multiple experiences with individual words. This research demonstrated that although new memories for individual items can be acquired rapidly, the process of discovering regularities across individual items to permit generalisation requires a period of overnight memory consolidation. The aim of the present research project is to discover whether sleep is the critical factor in the acquisition of general linguistic knowledge, and further, to describe the neural processes arising during sleep that facilitate this form of learning.
This proposal describes three work packages that combine methods at the leading edge of sleep science with our laboratory analogue of language learning to uncover how sleep impacts on the development of item-specific and general knowledge. In the first work package, we track the acquisition of item-specific and general knowledge when there is a delay between training and testing, and assess whether it matters if that delay consists of overnight sleep as opposed to daytime wake. In the second work package, we investigate how sleep deprivation before or after training impacts on the acquisition of item-specific and general linguistic knowledge. In the third work package, we use an olfactory cuing technique to reactivate memories of newly-learned information during sleep, and measure whether this reactivation enhances the acquisition of linguistic knowledge. We then take this experimental paradigm one step further to ask whether we can bias the course of long-term learning by selectively reactivating particular memories. In all experiments involving sleep, we use polysomnography to assess the importance of particular sleep stages or neural events during sleep for different forms of learning. We also assess learning in all experiments after one week to draw conclusions about the stability of new knowledge over the longer term.
International research has shown that the UK has one of the largest proportions of children who are sleep deprived, and that more than a quarter of the UK population gets on average less than five hours of sleep nightly. Given these statistics, it is of vital importance to understand what the consequences of poor sleep are for learning and memory. This research programme will address profound questions about how the brain continues to process new memories during sleep, how these sleep-related neural mechanisms shape the acquisition of long-term linguistic knowledge, and how sleep prepares the brain for new learning. Our findings will be transformative for theoretical models of learning, particularly as these apply to language, and will provide a range of new opportunities for creating substantive impacts within educational settings. We have developed a full programme of engagement with academic and non-academic stakeholders to realise this potential.
Recent research provides strong clues that sleep may play a vital role in the acquisition of general linguistic knowledge. In a previous ESRC project, we developed a laboratory analogue of language learning to track how general knowledge is built through multiple experiences with individual words. This research demonstrated that although new memories for individual items can be acquired rapidly, the process of discovering regularities across individual items to permit generalisation requires a period of overnight memory consolidation. The aim of the present research project is to discover whether sleep is the critical factor in the acquisition of general linguistic knowledge, and further, to describe the neural processes arising during sleep that facilitate this form of learning.
This proposal describes three work packages that combine methods at the leading edge of sleep science with our laboratory analogue of language learning to uncover how sleep impacts on the development of item-specific and general knowledge. In the first work package, we track the acquisition of item-specific and general knowledge when there is a delay between training and testing, and assess whether it matters if that delay consists of overnight sleep as opposed to daytime wake. In the second work package, we investigate how sleep deprivation before or after training impacts on the acquisition of item-specific and general linguistic knowledge. In the third work package, we use an olfactory cuing technique to reactivate memories of newly-learned information during sleep, and measure whether this reactivation enhances the acquisition of linguistic knowledge. We then take this experimental paradigm one step further to ask whether we can bias the course of long-term learning by selectively reactivating particular memories. In all experiments involving sleep, we use polysomnography to assess the importance of particular sleep stages or neural events during sleep for different forms of learning. We also assess learning in all experiments after one week to draw conclusions about the stability of new knowledge over the longer term.
International research has shown that the UK has one of the largest proportions of children who are sleep deprived, and that more than a quarter of the UK population gets on average less than five hours of sleep nightly. Given these statistics, it is of vital importance to understand what the consequences of poor sleep are for learning and memory. This research programme will address profound questions about how the brain continues to process new memories during sleep, how these sleep-related neural mechanisms shape the acquisition of long-term linguistic knowledge, and how sleep prepares the brain for new learning. Our findings will be transformative for theoretical models of learning, particularly as these apply to language, and will provide a range of new opportunities for creating substantive impacts within educational settings. We have developed a full programme of engagement with academic and non-academic stakeholders to realise this potential.
Planned Impact
This project will deliver substantial new knowledge of the mechanisms that underpin human learning, in particular with respect to the role of sleep in this process. There is substantial interest in sleep and its influence on different forms of learning in a range of communities. However, we believe that our work will have the most rapid and powerful impact on users within educational communities. These include parents, students, teachers, lecturers, head teachers, educational psychologists, SENCOs, speech and language therapists, users involved with educational charities, and policy-makers.
There are two reasons why these user groups should be interested in our work.
First, there are growing concerns about the impact of poor sleep on student performance. Research has already linked poor sleep to changes in children's mood, cognition, and behaviour; and there is evidence that lack of sleep is associated with poorer school performance (although much of this work is based on correlation; see [1] for review]). Our research will make a vital contribution to this body of knowledge, because it will test experimentally in a highly-controlled manner how different sleep manipulations enhance or impair the acquisition of different types of knowledge. We believe that our experiments using sleep deprivation in particular could yield quite dramatic impacts on learning. In the short term, such an outcome would be influential in raising awareness of the importance of sleep, and motivating behaviour change (e.g. schools issuing evidence-based guidance to parents regarding the importance of proper sleep for learning). In the mid- to longer-term, we believe that our findings could contribute to ongoing policy discussions pertaining to the most appropriate time to begin the school day.
Second, our experiments will feed into an enhanced understanding of human learning generally, which could have impacts on teaching practice. Our interaction with professionals involved in education suggests a considerable desire to develop evidence-based teaching practices that enhance learning in student populations. Our research will be of interest to these communities because it will reveal not just whether sleep is important for human learning, but why. Though we would not expect for our findings to yield widespread changes in teaching practice, we will use this project to continue to build closer networks between educational professionals and researchers with expertise in the science of learning. In doing so, it is important to emphasise that the benefits of these connections are two-way. Just as educational professionals may value contact with experts in the science of learning, our research is enhanced by understanding challenges in the delivery of effective teaching and learning, and discovering opportunities in which new research could have a substantial impact on activity in the classroom.
Our experience of building bridges with these users through other ESRC projects has been highly positive and effective. In order to continue to build these bridges and deliver tangible impacts, we have carefully planned a strategy for engaging with these users. This strategy includes presentations in fora that attract these users, exploiting and enhancing our connections within this area, capitalising on our South East Research Network for Schools to invite relevant practitioners to workshops focused on learning, and continuing to raise awareness through media contacts and by writing general interest articles in outlets with wide reach.
[1] Willingham, D. (2012). Are sleepy students learning? American Educator, winter, 2012-2013 edition.
There are two reasons why these user groups should be interested in our work.
First, there are growing concerns about the impact of poor sleep on student performance. Research has already linked poor sleep to changes in children's mood, cognition, and behaviour; and there is evidence that lack of sleep is associated with poorer school performance (although much of this work is based on correlation; see [1] for review]). Our research will make a vital contribution to this body of knowledge, because it will test experimentally in a highly-controlled manner how different sleep manipulations enhance or impair the acquisition of different types of knowledge. We believe that our experiments using sleep deprivation in particular could yield quite dramatic impacts on learning. In the short term, such an outcome would be influential in raising awareness of the importance of sleep, and motivating behaviour change (e.g. schools issuing evidence-based guidance to parents regarding the importance of proper sleep for learning). In the mid- to longer-term, we believe that our findings could contribute to ongoing policy discussions pertaining to the most appropriate time to begin the school day.
Second, our experiments will feed into an enhanced understanding of human learning generally, which could have impacts on teaching practice. Our interaction with professionals involved in education suggests a considerable desire to develop evidence-based teaching practices that enhance learning in student populations. Our research will be of interest to these communities because it will reveal not just whether sleep is important for human learning, but why. Though we would not expect for our findings to yield widespread changes in teaching practice, we will use this project to continue to build closer networks between educational professionals and researchers with expertise in the science of learning. In doing so, it is important to emphasise that the benefits of these connections are two-way. Just as educational professionals may value contact with experts in the science of learning, our research is enhanced by understanding challenges in the delivery of effective teaching and learning, and discovering opportunities in which new research could have a substantial impact on activity in the classroom.
Our experience of building bridges with these users through other ESRC projects has been highly positive and effective. In order to continue to build these bridges and deliver tangible impacts, we have carefully planned a strategy for engaging with these users. This strategy includes presentations in fora that attract these users, exploiting and enhancing our connections within this area, capitalising on our South East Research Network for Schools to invite relevant practitioners to workshops focused on learning, and continuing to raise awareness through media contacts and by writing general interest articles in outlets with wide reach.
[1] Willingham, D. (2012). Are sleepy students learning? American Educator, winter, 2012-2013 edition.
Publications
Castles A
(2018)
Ending the Reading Wars: Reading Acquisition From Novice to Expert.
in Psychological science in the public interest : a journal of the American Psychological Society
Cevoli B
(2022)
Prediction as a basis for skilled reading: insights from modern language models.
in Royal Society open science
Chang YN
(2020)
The relationships between oral language and reading instruction: Evidence from a computational model of reading.
in Cognitive psychology
D. P. Morgan
(2019)
Read me file for dataset from The impact of sleep on eyewitness identifications
Dawson N
(2021)
Finding the man amongst many: A developmental perspective on mechanisms of morphological decomposition.
in Cognition
Dawson N
(2021)
Bridging form and meaning: support from derivational suffixes in word learning
in Journal of Research in Reading
March J
(2023)
Is word learning capacity restored after a daytime nap?
in Cortex
Description | This research sought to discover how neural processes that arise during sleep impact on the learning of different forms of linguistic information. It combined methods at the leading edge of sleep science with a highly-controlled laboratory analogue of language learning that permits assessment of memory for specific items as well as general knowledge reflecting underlying structure across those items. The first work package investigated the impacts of total overnight sleep deprivation before and after learning on the acquisition of specific and general linguistic knowledge. To our surprise, although measures of vigilance showed dramatic effects of sleep deprivation through the night, we found no impacts of a full night of sleep deprivation on measures of memory for individual trained items, or on the induction of underlying regularities across the training sets, immediately or one week after training. These findings are important because they provide a well-powered test of the impact of sleep deprivation on learning and memory, and therefore suggest that previous claims about the impact of sleep on these cognitive functions may have been overstated. The observation that sleep deprivation did not influence learning and memory in our paradigm was surprising given previous literature. Thus, we conducted a quantitative meta-analysis of research on the impact of total sleep deprivation on learning and memory. This meta-analysis revealed a literature characterised by substantial challenges around adequate power and publication bias. Using advanced methods to estimate the impact of the publication bias, we still found a small effect of sleep deprivation across the whole literature, but it is not nearly as dramatic or profound as we originally anticipated, and few studies conducted across the whole literature had adequate power to detect the meta-analytic effect size. The next work package investigated whether we could bias participants' learning by reactivating particular memories during sleep. These memory reactivation effects have been reported previously for episodic memory, but not for the acquisition of general knowledge. Originally, we planned to implement a paradigm whereby memories were reactivated during particular stages of sleep using an olfactory cuing method. However, this proved more technically challenging than we planned, and raised health and safety issues that would have taken more time to resolve than we could afford. Thus, we implemented a more standard auditory cuing paradigm in which trained stimuli are played to participants overnight as they sleep. Once again, despite being well powered, we failed to observe any impacts of memory reactivation on episodic memory or general knowledge. This set of findings agrees with those observed in our sleep deprivation work package, but again, is not consistent with recent published work. Our analysis of the quality of these previous findings suggest that there may have been significant limitations around power. In a final work package, we tested the impact of normal overnight sleep on learning of novel words and generalisation of regular patterns. Once again, we observed no specific effect of sleep on these outcomes beyond time-of-day control conditions. Together, these studies raise questions about previous research reporting effects of sleep on language learning and generalisation. Our meta-analysis revealed that many of these studies had low power, and thus may have been susceptible to statistical error. However, it is unlikely that poor methodology explains all of the divergence. It is possible that sleep influences some domains of language learning more than others, or that there are individual differences in the extent to which sleep is necessary for learning and memory. These would be important questions for future research. |
Exploitation Route | The most important message of this project is probably around research methodologies, power, and Open Science. This is an area in which there have been rather sensational findings reported, but which we've found to be beset with problems of low power. Low power makes it difficult to find genuine findings, but also raises the risk of false positives. We believe that our own properly-powered studies and meta-analyses should lead to a re-evaluation of the role of sleep on behavioural indices of language learning. |
Sectors | Education |
Description | Our findings have had significant academic impact in refining understanding of the conditions under which sleep and sleep deprivation influence memory and learning. Studies in this area frequently have low participant numbers which makes them susceptible to Type I statistical errors. Our meta-analysis shows that while effects of sleep deprivation on memory and learning are significant, these tend to have small effect sizes and require substantial numbers of participants to observe. This conclusion is consistent with our empirical studies which show scarce evidence of these effects despite having unusually large participant numbers for this field. This grant has led to further funding from The Leverhulme Trust to assess the impacts of sleep behaviour on educational outcomes at a much larger scale. Our findings have had non-academic impact on public understanding of the wider relationship between sleep and cognition. We published three general interest articles in The Conversation related to this grant (see engagement activities), and these have had over 200,000 views. |
First Year Of Impact | 2022 |
Sector | Education |
Impact Types | Societal |
Description | How does poor sleep in adolescents impact learning and education? |
Amount | £242,129 (GBP) |
Organisation | The Leverhulme Trust |
Sector | Charity/Non Profit |
Country | United Kingdom |
Start | 01/2024 |
End | 12/2027 |
Description | Article in The Conversation about Auditory Processing during Sleep |
Form Of Engagement Activity | A magazine, newsletter or online publication |
Part Of Official Scheme? | No |
Geographic Reach | International |
Primary Audience | Public/other audiences |
Results and Impact | Tamminen, J. (2022). How the brain tunes in to unfamiliar voices while you're sleeping - and why it matters. The Conversation. https://theconversation.com/how-the-brain-tunes-in-to-unfamiliar-voices-while-youre-sleeping-and-why-it-matters-175018 |
Year(s) Of Engagement Activity | 2022 |
URL | http://theconversation.com/how-the-brain-tunes-in-to-unfamiliar-voices-while-youre-sleeping-and-why-... |
Description | Article in The Conversation about COVID and Sleep |
Form Of Engagement Activity | A magazine, newsletter or online publication |
Part Of Official Scheme? | No |
Geographic Reach | International |
Primary Audience | Public/other audiences |
Results and Impact | Tamminen, J., & Crowley, R. (2023). How COVID can disturb your sleep and dreams - and what could help. The Conversation. https://theconversation.com/how-covid-can-disturb-your-sleep-and-dreams-and-what-could-help-194798 |
Year(s) Of Engagement Activity | 2023 |
URL | http://theconversation.com/how-covid-can-disturb-your-sleep-and-dreams-and-what-could-help-194798 |
Description | Article in The Conversation about Sleep during Lockdown |
Form Of Engagement Activity | A magazine, newsletter or online publication |
Part Of Official Scheme? | No |
Geographic Reach | International |
Primary Audience | Public/other audiences |
Results and Impact | Tamminen, J., & Crowley, R. (2021). We're sleeping more in lockdown, but the quality is worse. The Conversation. https://theconversation.com/were-sleeping-more-in-lockdown-but-the-quality-is-worse-155797 |
Year(s) Of Engagement Activity | 2021 |
Description | Engagement with Balfour Beaty |
Form Of Engagement Activity | A talk or presentation |
Part Of Official Scheme? | No |
Geographic Reach | Regional |
Primary Audience | Industry/Business |
Results and Impact | Discussed impact of sleep restriction on performance, particularly in relation to shift work |
Year(s) Of Engagement Activity | 2021 |
Description | Interview with Digital Trends |
Form Of Engagement Activity | A broadcast e.g. TV/radio/film/podcast (other than news/press) |
Part Of Official Scheme? | No |
Geographic Reach | National |
Primary Audience | Industry/Business |
Results and Impact | Discussed how sleep-dependent consolidation is linked to language learning and its implications for education. |
Year(s) Of Engagement Activity | 2021 |
Description | Interview with Sputnik Radio |
Form Of Engagement Activity | A broadcast e.g. TV/radio/film/podcast (other than news/press) |
Part Of Official Scheme? | No |
Geographic Reach | International |
Primary Audience | Public/other audiences |
Results and Impact | Discussed sleep changes during COVID-19 lockdowns, and relation of sleep restriction to achievement and learning. |
Year(s) Of Engagement Activity | 2021 |
Description | Presentation to teachers regarding sleep changes in adolescence |
Form Of Engagement Activity | Participation in an activity, workshop or similar |
Part Of Official Scheme? | No |
Geographic Reach | Regional |
Primary Audience | Professional Practitioners |
Results and Impact | Presentation to SouthEast Network for Schools to ~100 teachers regarding sleep changes in adolescence and implications for learning |
Year(s) Of Engagement Activity | 2018 |
Description | Sleep Research Demonstration |
Form Of Engagement Activity | Participation in an open day or visit at my research institution |
Part Of Official Scheme? | No |
Geographic Reach | Local |
Primary Audience | Public/other audiences |
Results and Impact | Demonstration of sleep research for Royal Holloway Science Open Day |
Year(s) Of Engagement Activity | 2019 |