Chaos has long been a topic for discussion in the hypertext community. In one of the core founding documents of hypertext. In Halasz’ ‘Seven Issues’ paper, he characterises hypertext as a way of organising the chaos in our heads [14]. Conversely, Marchionini argued that the introduction of hypertext to the classroom would bring opportunity, but also chaos, linking the two inextricably [26]. So which is true? Does hypertext add to the chaos, or does it help us reduce the chaos in our heads? Over thirty years have passed since those papers were written, and in that time we have seen an explosion in both the use of hypertext-based systems (in the form of the world wide web), and in algorithms and tools to manage the chaos and make it tractable. Some of the most cutting edge research in the Hypertext conference since that time has been in creating chaos-generating interfaces, such as Kerne's generative hypermedia interface [18].

We are facing novel information challenges: search is becoming increasingly focused on providing the one right answer, to the exclusion of an information seeking process which might expose users to a wide range of information and choices along the way [41]. Meanwhile, social media has become a wild west of confounding information, misinformation, and contentious viewpoints [12; 42; 46]. In response to this, social media users might engage with views other than their own, but they also often mute, unfriend, or block people espousing different views [31; 40]. Physical information spaces, known for supporting creativity and serendipity, are disappearing, and online interfaces to replace them have been slow to emerge [1; 24; 28; 43; 45]. The thread that links all of these challenges is information diversity, or, from a user perspective, presenting just the right amount of chaos in any given situation. In this paper we examine the ways in which the web has promoted chaos-limiting strategies, then the benefits of chaos before setting out a pro-chaos research agenda for the hypertext community.

Two common chaos limiting strategies exist on the web: one technical (search), the other social (creating echo chambers).

Search engines are designed specifically to address an abundance of information, but also to meet tightly specified information needs. Early search engines were assessed for their ability to provide the exact set of documents that was relevant to a searcher's query in descending order of relevance likelihood [2]. Increasingly, though, with the emergence of conversational search and machine learning based retrieval models, the ideal goal is one single piece of information that will meet a searcher's needs [38]. Search engines, far from representing the underlying documents and allowing users to assess and integrate them, are now constructing composite documents that answer a query, limiting the options for searchers to engage with original documents. [41] Arguably this has been driven by user demand: many users click on the first document in a search [17], and a dearth of relevant documents high up in a search result list creates a sense of ‘too many’ documents [19]. In an increasingly adversarial information environment though, seeing only a single perspective on an information limits opportunities for learning, civic engagement, and perhaps democracy itself [15; 16].

In contrast to search, social media is a platform of seemingly high democracy: everyone is able to say something (however ill informed or inexpert), and to some it seems they are all saying it at once [42]. The context collapse on social media (where different parts of one's life all meet in the same digital place) is often overwhelming [7; 9], and social media algorithms have recently been revealed to have been driving engagement using outrage [3; 32]. Between the high levels of visual and topical diversity, and the repeated experience of encountering information with which they disagree drives some users to constrain their social media channels [39]. They do this in a variety of ways, most notably by muting, blocking or unfriending people who continually express views with which they disagree [31; 40]. This reduces the viewpoint diversity in their feeds, calming one source of the chaos [4; 12]. Another approach is to ‘game’ recommendation algorithms on social media by deliberately not clicking on or engaging with disagreeable content [31], notably on TikTok—not engaging with videos ‘tells’ the algorithm not to offer more like that.

Clearly, there is a long history of chaos limiting online, and deep social engagement with it; this aligns with Halasz’ view of chaos as something hypermedia can help to manage [14]. Even serendipity, which is usually perceived as delightful, can be distracting and a negative intrusion on a pre-determined task [44], and there are benefits of echo chambers for refining complex ideas [9; 15]. Why, then, are we arguing for chaos?

McKay et al have made a comprehensive case for the power of chaos in serendipity and invention: they argue that the right amount of chaos prompts new ideas by bringing previously-unconsidered concepts together in ways that are exciting and delightful rather than overwhelming and exhausting [29]. They note that the chaos in such situations cannot be unconstrained, though, and point to curating social media feeds as one way of limiting or bounding that chaos. Given the global challenges we are facing that will need high levels of innovation to solve, creativity and serendipity are not just nice to have, they are a global imperative [8; 35; 37].

Alongside creativity and serendipity, though, is the issue of misinformation and negative echo chambers. We know that people will share misinformation they agree with even if they know it to be false [11]. We also know that serendipitous information encounters with information we may find disagreeable initially can lead to empathy with those who hold the view, or a complete change of view [30]. For these encounters to happen, though, people must be exposed to a diverse range of views [15; 36; 40; 47]. Some people deliberately expose themselves to a wide range of views, generating their own chaos [30]. For this to only happen where users seek it, though, doesn't meet the diversity principle legal scholars have determined is the underpinning of democracy [16]. We must all be exposed to a range of views to be informed, and we must be informed to participate in democracy. While it has been argued that this is the purview of search engines, as we saw above search engines are decreasing, rather than increasing the amount of information shown to users. It is also likely that the format, timing, language and source of information will affect reader receptivity [6; 10; 34], so it is not as simple as presenting more diverse search results. Thus we argue not for unlimited chaos, where outright rejection of alternative ideals may leave people more polarised, but instead precision chaos, where alternative ideas are most likely to be considered, if ultimately rejected. We are not arguing to manipulate users, but instead to give them choices. The question is how to gain the creative and democratic benefits of chaos, without the negative outcomes?

It is clear that, like any force of nature, informational chaos is generative and valuable in the right amount [24; 25; 33], and destructive and overwhelming where it is too much [5]. The challenge is to determine which is which. We need information to inspire creativity and challenge misinformation-based echo chambers, but we also need space to discuss the finer points of an idea without being drowned out by opposing viewpoints. In this section we build on the research agenda set out for the information interaction community in [29], incorporating the challenges and benefits of filter bubbles and echo chambers, and focusing specifically on the hypertext community. We talk first about building an understanding of successful chaos, second about creating algorithms to support that chaos, and finally about interfaces to allow users to manipulate the amount of chaos to which they are exposed in a given moment.

To generate successful chaotic interfaces, we must first understand successful chaos. Some of this work is already done: Makri gives a clear account of the process [22; 23] and necessary preconditions [24] for serendipity, for example. There is still much we do not know, however. Social media is both a place that can become polarised, and a place people can feel overwhelmed and thus limit the chaos by closing their social circles [40]. It can be the beginning of positive, pro-social view changes, and it can lead people down a rabbit hole of misinformation. The ability to narrow down our social groups can allow us to refine our ideas [9], or it can prevent us from meeting new ones [12; 13]. Turning to other interfaces, getting lost on Wikipedia [27] or in a music playlist [20] can be a transporting experience, or it can be a distraction from a real task [44]. To ensure hypertext interfaces facilitate pro-social change both on social media and on the web more generally, we need to understand and define successful chaos. Clearly, the kind of chaos that prompts new ideas that result in new scientific discoveries is successful, but what about the chaos that makes someone uncomfortable, but also reduces polarisation by breaching an echo chamber? This is a matter for the Hypertext community, and for users to decide. To understand the qualities of successful chaos will require both qualitative and quantitative analyses. We need to look at individual examples of successful chaos to understand the user experience. We also need to analyse quantitatively the types of features—timing, similarity to existing views, linguistic features—that characterise successful chaos; and the most likely environment for this is social media. The Hypertext community has the research background and skills to do this work.

Once we have characterised successful chaos, we will need algorithms to support it These algorithms will need to supply the right information in the right context at the right time for success. McKay noted in 2020 that algorithms for chaotic information interfaces would ‘present a non-deterministic sample of an underlying collection to a user, where there is no clear principle on which the sample could be selected’ [29]. We argue that this definition is necessary, but not sufficient. Merely having a non-deterministic sample will not produce successful serendipity, creativity, filter bubble breaking, though. Instead we need to derive a set of likely-acceptable, or at least likely-inoffensive, items to present [10; 21]. Once this set is derived, we need to determine how (in what format, and with what contextualisation), and when to present a non-deterministic subset of our sample to the user: the right information at the right time is a key feature of serendipity [24]. The Hypertext community, with its long history of making interesting links between content, and social media analysis and assessment, is ideally placed to do this work.

We know that people have different appetites for serendipity, or information they disagree with, depending on their mood, the type of information available and a rage of other factors. McKay pointed out in 2020, that to be useful chaos needs bounding or restriction [29]. The mechanisms proposed in that paper, though, such as using fixed library interfaces, or curating one's social media friends’ list to remove disagreeable information from social media feeds are relatively static, not changing greatly over time. This does not reflect users’ needs or tolerance for chaos, which both change and are influenced by their current context. The final piece of the puzzle the hypertext community can offer is developing interface mechanisms that allow users to provide feedback about the current chaos level, asking for more when the situation warrants it, or reducing it when they are overloaded. Again, the Hypertext community, with their long history of developing complex online user interactions with information, are ideally placed to meet this need.

Ted Nelson has long quoted as saying that “the objective of Hypertext research is to save the planet”. In this paper we outline one way in which that might be true. To solve wicked problems, shatter echo chambers, reduce polarisation and support social cohesion and democracy, people need to be exposed to a range of viewpoints and ideas. Simply throwing ideas at people, though, is unproductive, they become overloaded and overwhelmed, rather than curious and engaged. In this paper we set out a research agenda for Hypertext that encourages just the right amount of chaos in any given situation, so that we might all become more comfortable with chaos.

Andre, P., schraefel, m.c., Teevan, J., and Dumais, S.T., 2009. Discovery is never by chance: designing for (un)serendipity. In Proc. C&C 09 (Berkeley, California, USA), ACM, 1640279, 305-314. DOI= http://dx.doi.org/10.1145/1640233.1640279.

Baeza-Yates, R. and Ribeiro-Neto, B., 1999. Modern information retrieval. ACM press New York.

Bakir, V. and McStay, A., 2018. Fake News and The Economy of Emotions. Digital Journalism 6, 2, 154-175. DOI= http://dx.doi.org/10.1080/21670811.2017.1345645.

Bakshy, E., Messing, S., and Adamic, L.A., 2015. Exposure to ideologically diverse news and opinion on Facebook. Science 348, 6239, 1130-1132. DOI= http://dx.doi.org/doi:10.1126/science.aaa1160.

Bawden, D. and Robinson, L., 2009. The dark side of information: overload, anxiety and other paradoxes and pathologies. J Inf Sci 35, 2, 180-191. DOI= http://dx.doi.org/10.1177/0165551508095781.

Benegal, S.D. and Scruggs, L.A., 2018. Correcting misinformation about climate change: the impact of partisanship in an experimental setting. Climactic Change 148, 1, 61-80. DOI= http://dx.doi.org/10.1007/s10584-018-2192-4.

Boyd, D., 2014. It's complicated: The social lives of networked teens. Yale University Press.

Brown, V.A., Harris, J.A., and Russell, J.Y., 2010. Tackling wicked problems through the transdisciplinary imagination. Earthscan, London, UK.

Bruns, A., 2019. Filter bubble. Internet Policy Review 8, 4. DOI= http://dx.doi.org/10.14763/2019.4.1426.

Chan, M.-p.S., Jones, C.R., Hall Jamieson, K., and Albarracín, D., 2017. Debunking: A Meta-Analysis of the Psychological Efficacy of Messages Countering Misinformation. Psyc. Sci 28, 11, 1531-1546. DOI= http://dx.doi.org/10.1177/0956797617714579.

Chen, X., Sin, S.-C.J., Theng, Y.-L., and Lee, C.S., 2015. Why Do Social Media Users Share Misinformation? In Proc. JCDL 15 (Knoxville, Tennessee, USA), Association for Computing Machinery, 111–114. DOI= http://dx.doi.org/10.1145/2756406.2756941.

Colleoni, E., Rozza, A., and Arvidsson, A., 2014. Echo chamber or public sphere? Predicting political orientation and measuring political homophily in Twitter using big data. J Comm 64, 2, 317-332.

Fletcher, R. and Nielsen, R.K., 2017. Are people incidentally exposed to news on social media? A comparative analysis. New Media & Society 20, 7, 2450-2468. DOI= http://dx.doi.org/10.1177/1461444817724170.

Halasz, F.G., 1988. Reflections on NoteCards: seven issues for the next generation of hypermedia systems. CACM 31, 7, 836–852. DOI= http://dx.doi.org/10.1145/48511.48514.

Helberger, N., Karppinen, K., and D'Acunto, L., 2018. Exposure diversity as a design principle for recommender systems. Inf. Comm & Soc. 21, 2, 191-207. DOI= http://dx.doi.org/10.1080/1369118X.2016.1271900.

Helberger, N., Kleinen-von Könisglöw, K., and van der Noll, R., 2015. Regulating the new information intermediaries as gatekeepers of information diversity. Information and Media 17, 6, 50-71.

Jansen, B.J., Spink, A., Bateman, J., and Saracevic, T., 1998. Real life information retrieval: a study of user queries on the Web. SIGIR Forum 32, 1, 5–17. DOI= http://dx.doi.org/10.1145/281250.281253.

Khandelwal, M., Kerne, A., and Mistrot, J.M., 2004. Manipulating history in generative hypermedia. In Proc. HT 04 (Santa Cruz, CA, USA), Association for Computing Machinery, 139–140. DOI= http://dx.doi.org/10.1145/1012807.1012848.

Khoo, M. and Hall, C., 2012. What would 'Google' Do? Users' Mental Models of a Digital Library Search Engine. In Proc. TPDL 12 (Paphos, Cyprus), Springer, Berlin, Germany, 1-12. DOI= http://dx.doi.org/10.1007/978-3-642-33290-61.

Leong, T.W., Vetere, F., and Howard, S., 2005. The serendipity shuffle. In Proc. OzCHI 05 (Canberra, Australia), ACM, New York, 1-4.

Lewandowsky, S., Ecker, U.K.H., Seifert, C.M., Schwarz, N., and Cook, J., 2012. Misinformation and Its Correction: Continued Influence and Successful Debiasing. Psyc Sci in Public Interest 13, 3, 106-131. DOI= http://dx.doi.org/10.1177/1529100612451018.

Makri, S. and Blandford, A., 2012. Coming across information serendipitously: Part 1 – A process model. J Doc 68, 5, 685-704. DOI= http://dx.doi.org/10.1108/00220411211256030.

Makri, S. and Blandford, A., 2012. Coming across information serendipitously: Part 2 - A classification framework. J Doc 68, 5, 706-724. DOI= http://dx.doi.org/10.1108/00220411211256049.

Makri, S., Blandford, A., Woods, M., Sharples, S., and Maxwell, D., 2014. “Making my own luck”: Serendipity strategies and how to support them in digital information environments. JASIST 65, 11, 2179-2194. DOI= http://dx.doi.org/10.1002/asi.23200.

Makri, S., Ravem, M., and McKay, D., 2017. After serendipity strikes: Creating value from encountered information. ASIST Proceedings 54, 1, 279-288. DOI= http://dx.doi.org/10.1002/pra2.2017.14505401031.

Marchionini, G., 1988. Hypermedia and Learning: Freedom and Chaos. Educ. Tech. 28, 11, 8-12.

McCay-Peet, L. and Toms, E.G., 2011. The serendipity quotient. ASIS&T Proceedings 48, 1, 1-4. DOI=https://doi.org/10.1002/meet.2011.14504801236.

McKay, D., Buchanan, G., and Chang, S., 2018. It ain't what you do, it's the way that you do it: Design guidelines to better support online browsing. ASIS&T Proceedings 55, 1, 347-356. DOI=http://dx.doi.org/10.1002/pra2.2018.14505501038.

McKay, D., Makri, S., Chang, S., and Buchanan, G., 2020. On Birthing Dancing Stars: The Need for Bounded Chaos in Information Interaction. In Proc. CHIIR 2020 (Vancouver BC, Canada), Association for Computing Machinery, 292–302. DOI=http://dx.doi.org/10.1145/3343413.3377983.

McKay, D., Makri, S., Gutierrez-Lopez, M., MacFarlane, A., Missaoui, S., Porlezza, C., and Cooper, G., 2020. We are the Change that we Seek: Information Interactions During a Change of Viewpoint. In Proc. CHIIR 20 (Vancouver BC, Canada), Association for Computing Machinery, 173–182. DOI= http://dx.doi.org/10.1145/3343413.3377975.

McKay, D., Owyong, K., Makri, S., and Lopez, M.G., 2022. Turn and Face the Strange: Investigating Filter Bubble Bursting Information Interactions. In Proc. CHIIR 22 (Regensburg, Germany), Association for Computing Machinery, 233–242. DOI= http://dx.doi.org/10.1145/3498366.3505822.

Mochon, D. and A Schwartz, J., 2019. Outrage drives facebook users to engage with ideology inconsistent political content. ACR North American Advances.

Muller, M., 2014. Curiosity, Creativity, and Surprise as Analytic Tools: Grounded Theory Method. In Ways of Knowing in HCI, J.S. OLSON and W.A. KELLOGG Eds. Springer New York, 25-48. DOI= http://dx.doi.org/10.1007/978-1-4939-0378-82.

Nyhan, B. and Reifler, J., 2010. When Corrections Fail: The Persistence of Political Misperceptions. Political Behaviour 32, 2, 303-330. DOI= http://dx.doi.org/10.1007/s11109-010-9112-2.

Peters, B.G., 2017. What is so wicked about wicked problems? A conceptual analysis and a research program. Policy and Society 36, 3, 385-396. DOI= http://dx.doi.org/10.1080/14494035.2017.1361633.

Reviglio, U., 2017. Serendipity by design? How to turn from diversity exposure to diversity experience to face filter bubbles in social media. In Proc. INSCI 17 (Thessaloniki, Greece), Springer, Berlin, 281-300.

Runco, M.A., 2010. Divergent thinking, creativity, and ideation. In The Cambridge handbook of creativity, 446.

Sanderson, M., 2010. Test collection based evaluation of information retrieval systems. In Foundations and Trends in Information Retrieval Now Publishing, Boston, MA, 247-375.

Sano, F., 2018. Designing Outrage, Programming Discord: A Critical Interface Analysis of Facebook as a Campaign Technology. Tech Comm 65, 4, 387-410.

Seargeant, P. and Tagg, C., 2019. Social media and the future of open debate: A user-oriented approach to Facebook's filter bubble conundrum. Doiscourse Context and Media 27, 41-48. DOI= http://dx.doi.org/10.1016/j.dcm.2018.03.005.

Shah, C. and Bender, E.M., 2022. Situating Search. In Proc. CHIIR 22 (Regensburg, Germany), Association for Computing Machinery, 221–232. DOI= http://dx.doi.org/10.1145/3498366.3505816.

Sunstein, C., 2018. #Republic: Divided Democracy in an Age of Social Media. Princeton University Press.

Thudt, A., Hinrichs, U., and Carpendale, S., 2012. The bohemian bookshelf: supporting serendipitous book discoveries through information visualization. In Proc. CHI 12 (Austin, TX), ACM, 2208607, 1461-1470. DOI= http://dx.doi.org/10.1145/2207676.2208607.

Waugh, S., McKay, D., and Makri, S., 2017. 'Too Much Serendipity': The Tension between Information Seeking and Encountering at the Library Shelves. In Proc. CHIIR 17 (Oslo, Norway), ACM, New York NY, 277-280. DOI= http://dx.doi.org/10.1145/3020165.3022132.

Whitelaw, M., 2015. Generous Interfaces for Digital Cultural Collections. DHQ: Digital Humanities Quarterly 9, 1, 2-2.

Wu, L., Morstatter, F., Carley, K.M., and Liu, H., 2019. Misinformation in Social Media: Definition, Manipulation, and Detection. SIGKDD Explor. Newsl. 21, 2, 80–90. DOI= http://dx.doi.org/10.1145/3373464.3373475.

Yom-Tov, E., Dumais, S., and Guo, Q., 2013. Promoting Civil Discourse Through Search Engine Diversity. Soc Sci Comp Rev 32, 2, 145-154. DOI= http://dx.doi.org/10.1177/0894439313506838.

Permission to make digital or hard copies of part or all of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page. Copyrights for third-party components of this work must be honored. For all other uses, contact the owner/author(s).

HT '22, June 28–July 01, 2022, Barcelona, Spain

© 2022 Copyright held by the owner/author(s). ACM ISBN 978-1-4503-9233-4/22/07. DOI: https://doi.org/10.1145/3511095.3536371