Abstract
In this paper, indigenous knowledge is viewed as a form of collective intelligence that may inform situational action. We consider knowledge associated with the use of fire by Indigenous Australians as a land management tool that may help moderate the impact of climate change. Scientific study of the foundation of such knowledge has enhanced our understanding of it, and combination with new technology tools can enhance traditional practices. We draw on the concept of a ‘collective intelligence genome’ identified in the literature to examine aspects of indigenous knowledge capture and application, and an enhanced representation of this concept is presented. We observe knowledge application draws on a set of rules, selection from which is dependent on the state of fauna and weather in a particular microcosm. A representation of multiple if-then decision-making is presented as an example of genome component application.
Introduction
The authors have a background interest in exploring regional dynamics supporting the achievement of UN sustainable development goals. Following the lead of others, we suggest underlying dynamic elements may be viewed through a collective intelligence lens (Fleming et al., 2007; Gray et al., 2020; Iandoli et al., 2007; Komninos, 2004; Salminen and Harmaakorpi, 2012). The theme we follow in this paper is the pursuit of United Nations Sustainable Development Goal SDG 13 – ‘Take urgent action to combat climate change and its impacts’ (https://sdgs.un.org/goals/goal13). Leach et al. (2012) contend that innovation in this context ‘requires a radically new approach to innovation, one that gives far greater recognition and power to grassroots actors and processes, involving them within an inclusive, multi-scale innovation’. Some authors (e.g. Moallemi et al., 2020) contend that transdisciplinary innovation and the incorporation of indigenous knowledge at a local scale are needed to achieve SDGs. Some researchers draw specific attention to the benefits of inclusive innovation in the pursuit of SDGs (e.g. Kalkanci et al., 2019; Senanayake, 2006). Sen (2005) argues that the need to bring global knowledge to developing countries for their benefit should be moderated by reciprocal learning about complementary indigenous knowledge which has accumulated over hundreds or thousands of years. Working across community and professional boundaries from multiple viewpoints is a recurring theme (e.g. Moallemi et al., 2020). In an indigenous community engagement context, this has been characterised as ‘two-way seeing’ (e.g. Smith, 1999).
Some researchers (e.g. Aminpour et al., 2020; Malone and Klein, 2007) had argued that collective intelligence tools could be utilised to confront climate change issues, and we followed this line of enquiry. Our research question was: how might indigenous knowledge that could support the pursuit of SDGs for mutual benefit be represented from a collective intelligence perspective? We drew on an Australian case example intended to mitigate the impact of more frequent wildfires that are one effect of climate change. Indigenous knowledge was combined with institutional knowledge to better prepare for potential wildfires. The paper starts with a review of the literature related to indigenous knowledge capture and utilisation and collective intelligence concepts. An inductive research strategy drawing on case material leads to a proposed representation of higher order interactions between observations, collective intelligence rules and collective intelligence resources.
Literature review
We briefly explore emergent themes in two bodies of literature: one related to indigenous knowledge capture and utilisation, and the other considering collective intelligence concepts.
Indigenous knowledge
Some issues associated with the sharing and utilisation of indigenous knowledge: Examples from the literature.
Table 1 indicates an interplay between government policy, the emergence of technologies that may support the preservation and utilisation of indigenous knowledge and the need to appreciate indigenous ‘ways of knowing’ that are connected to the land (e.g. Barnhardt and Kawagley, 2005). Christianson (2015) reviewed social science research on indigenous fire management practices in Australia, Canada and the USA post-2000. She described some potential communication and protocol obstacles to doing research with indigenous communities that required the establishment of common understandings and sensitive relationships. Verran and Christie (2013) described an approach to ‘doing difference’ separately before jointly forming concepts and confronting ‘ontic discomfort’ (which may subsequently lead to an innovative solution). Reid et al. (2021) promoted the idea of ‘two-eyed seeing as “a pathway for plural existence where time-tested indigenous knowledge systems can be paired with, not subsumed by Western scientific insights for an equitable and sustainable future”’.
Indigenous knowledge evolved over long periods of time from the direct observations of multiple actors. It was generated and utilised by the community for the community. Story-telling and experiential learning mechanisms practiced by trusted community members supported long-term retention reinforced by use in practice. If this knowledge is utilised by others in a post-colonial setting without providing benefit to the indigenous community there may be concerns about its correct interpretation and issues of mutual trust (e.g. Smith, 1999). This situation is not unique to indigenous communities. Lefebvre and Redien-Collot (2013) studied experiential learning associated with nascent entrepreneur mentoring. They noted the importance of inter-personal communication strategies and the establishment of mutual trust. Copeland and Moore (2018) identified four trust dimensions as being imperative foundations in implementing community digital media interventions for the common good: legitimacy, authenticity, synergy and commons. The point to be made here is that some of the concerns of indigenous communities of practice are also concerns of other communities, and such comparisons may be a topic for further research.
Collective intelligence and shared experience
The process of mobilising collective intelligence draws on the shared experience of people who have directly or indirectly been engaged in past events and may inform current (or future) action or create an improved knowledge base. It has been suggested that the outcome may be superior to that suggested by an experienced individual (e.g. Surowiecki, 2005 – The Wisdom of Crowds, Barnett et al., 2019). IT tools may be used to facilitate knowledge accumulation and sharing (e.g. the Wikipedia) to facilitate decision-making processes, drawing on large data sets (e.g. Hunter, 2005; Matarić, 2000; Mulgan, 2018). Some favour a hybrid approach that preserves societal context (e.g. Peeters et al., 2021). There may be subjective and objective perceptions of input, process and output quality to be considered, along with trust in the contributors (e.g. Copeland and De Moor, 2018) and matters of absorptive capacity (an ability to interpret and utilise associated knowledge – e.g. Blohm et al., 2011). The following outlines approaches to characterising the mechanisms involved.
The collective intelligence genome
Characterisation of a collective intelligence genome (adapted from Wise et al., 2012).
Drawing on a literature review, Salminen and Harmaakorpi (2012) suggested that human collective intelligence processes may be considered at three levels: At a micro-level, where a combination of psychological, behavioural elements and trust influence engagement in a social setting. At a macro-level, where ‘how’ processes similar to those in Table 1 are observed. An emergence level between micro and macro-levels supports adaptivity, interaction and rules followed at a local level to reinforce practice that emerges from macro-level processes. More recently, Suran et al. (2020) reviewed research on Collective Intelligence frameworks, providing more sub-element detail associated with the Genome shown in Table 2 and included a view of ways the ‘who’ and ‘how’ elements may interact in supporting the evolution of a particular instance. This highlights the existence of cumulative learning at all levels.
Collective intelligence – knowledge accumulation
Knowledge as a resource
Boder (2007) had argued that capturing collective intelligence was an important aspect of knowledge management, creating a resource that may support the maintenance of a competitive advantage. He suggested there were three building blocks of collective intelligence in an enterprise context: • Goals drawing on strategic market knowledge • Competencies drawing on the enterprise domain-specific knowledge and • Mechanisms, drawing on cultural norms
Boder (2007) also suggested firstly, that the extent to which knowledge captured in one context could be used in a different context, and secondly that a combination of internal and external collaboration supported the establishment of richer knowledge sets. As noted earlier, potential quality assurance matters may have to be addressed (e.g. Lichtenstein and Parker, 2009).
Knowledge as rules
Some researchers see ‘knowledge as rules that reduce environmental uncertainty through connections between ideas and facts’ (Tywoniak, 2007).
The notion of following and adapting rules resonates with structuration theory (Giddens 1984) that evolved from observations about the workings of social groups or larger communities. Social coherence is maintained by knowledgeable agents working with rules of signification (sensemaking), legitimation (accepted, endorsed practice) and domination (permission to use or allocate resources). Satisfactory outcomes reinforce the validity of the rules (called the duality of structure). Unsatisfactory outcomes indicate a need for change. Giddens (1985) also indicated that acceptable rules may be time and place dependent. This is consistent with the observations of Matarić (2000) that ‘intelligent behaviour is inextricably tied to its cultural context and … intelligent collective behaviour in a decentralised system results from local interactions based on simple rules’.
Utilising collective intelligence
Collective intelligence may be utilised as rules in decision-making, for example, in voting, or as a knowledge resource, for example, in problem-solving. Each of these aspects is discussed in the following.
The decision-making context
IBM knowledge management researchers Kutz and Snowden (2003) suggest that traditional decision support and strategy assumes there is an underlying order that supports rational choice. They evolved a sense-making concept (described as the Cynefin framework) that also considered appropriate action in unordered environments. Two ordered-environment and two unordered-environments were characterised as follows: • In a known, ordered environment, it is suggested that repeatability allows for predictive models to be established and best practice identified. Appropriate actions are sense–categorise–respond. • In an ordered environment where cause and effect are knowable, but not immediately evident (or may be known by only a few people), appropriate actions are sense–analyse–respond. • In an unordered environment where complex relationships and interactions cause unexpected outcomes, appropriate actions are seen as probe–sense–respond. In this environment, there may be a string of cause and effect relationships between the agents, and both the number of agents and the number of relationships make categorisation or analysis difficult. Kutz and Snowden suggested that emergent patterns may be perceived but not predicted; a phenomenon they called ‘retrospective coherence’, and see this as the domain of complexity theory. • In an unordered environment that seems chaotic, it is suggested an appropriate response is act–sense–respond. Kutz and Snowden have observed this situation where multiple decision-makers observe the same phenomenon from different points of view, and ‘those most comfortable with stable order seek to create or enforce rules; experts seek to conduct research and accumulate data; politicians seek to increase the number and range of their contacts; and finally, the dictators, eager to take advantage of a chaotic situation, seek absolute control’. They suggest that collaborating to reach consensus on a series of small actions can progress the situation.
Turpin and Marais (2004) noted the importance of sensitivity to the decision-making context that included consideration of socio-political factors and the actors involved. Others noted the influence of temporal factors – decision time horizons, timeliness, and time stress on decision-making (e.g. Ariely and Zakay, 2001; Hogenboom et al., 2021).
Decision-making actors
Suran et al. (2020) noted two aspects of actor interaction in their studies of collective intelligence projects: (a) establishing trust and respect and (b) the nature of interactions in knowledge diffusion activities characterised in the SECI model (Nonaka et al., 2000). In collaborative ventures, goal congruence and the maintenance of trust is supported by the enactment of three roles: (1) a communicator role helps share available information, (2) a coordinator role helps organise tasks, and (3) a collaborator role helps maintain relationships and resolve tension (Zhou, 2001). King and Lakhani (2011) noted the contingent effect of actor absorptive capacity in an open innovation context where internal and external experiences were combined.
Drawing on knowledge sets
We suggest that just as snippets of knowledge are drawn on in a collective intelligence context, they can be assembled in a variety of ways in the application of this collection. Some examples follow. Experience-based stories have been used as a resource in the utilisation of case-based reasoning tools. Bosch et al. (1997) described the use of case-based reasoning technologies to build on a combination of local and scientific knowledge in resolving rangeland management problems. Bergmann et al. (2005) defined a case used in case-based reasoning as a contextualised piece of experience that may take different forms. Whatever the form, it was recommended that cases include a problem situation description (who, where and when), the proposed solution (what and how), the outcome the consequences of its application and the extent to which the original problem was addressed (why the solution made sense).
Roos (2013) had studied the stories of indigenous South-Eastern Australians that reflected climate change experience over thousands of years. He suggested there was a correlation between the structure and use of these stories and pattern language. Originally introduced as a way of viewing building architecture as a selective configuration of generic elements, the concept has been applied in other fields, for example, information systems (Alexander, 1999; Beedle et al., 1999). More recently, pattern language has been used as a collective intelligence tool associated with ‘living lab’ activities (Akasaka et al., 2020; Iba et al., 2016). In the latter application, assembled groups of experienced practitioners and users share their experiences which are then organised into sets of patterns. A form of template having the following parts is used: establish context (where and when), the problem (what), the solution (how), the solution application (by who and why) plus a real-world example (associated story).
In summary
Indigenous people have been utilising collective intelligence mechanisms for hundreds or thousands of years and have accumulated knowledge as a resource and knowledge as rules. They also have protocols defining who does what in sharing and drawing on this knowledge. But there are concerns this knowledge and ways of continuously revitalising it may be lost if traditional knowledge diffusion mechanisms are disrupted. In addition, how this knowledge can be interpreted and integrated with contemporary scientific knowledge for mutual benefit is being explored.
A ‘Collective Intelligence Genome’ concept considers matters of what (the focus of the activity), who (should be involved), why (the motivation) and how (mechanisms). These mechanisms included developing a common viewpoint on valid knowledge and decisions via analysis processes (analysing, categorising, averaging, voting or consensus), collecting information to inform decision-making in a complex environment (act, sense and probe). This set was considered to be appropriate independent of the application domain. However, drawing on the foregoing brief discussion about knowledge as rules and decision-making context, we add consideration of where (external context) and when (timeliness) to this set, factors that are seen as particularly relevant in an indigenous knowledge context (Karsten and Illa, 2005; Ode and Ayavoo, 2020).
Research approach
Our research question was: how might indigenous knowledge that could support the pursuit of SDGs be represented for mutual benefit from a collective intelligence perspective in the context of SDG 13 – ‘Take urgent action to combat climate change and its impacts’. Yin (2014) suggests that a case study method may be appropriate when investigating a ‘how’ question, particularly where the researcher has little or no control over behavioural events. The research question is very broad, and the literature survey suggests there are many factors to consider in capturing and utilising indigenous knowledge (Table 1). In this paper, we focus on one aspect of climate change that has significantly impacted our region – an increasing severity and frequency of wildfires. Australian aborigines and Torres Strait islanders have used fire as a land management tool for thousands of years, adopting a technique called ‘cultural (or cool) burning’, and it has been suggested this practice may help reduce the impact of wildfires (e.g. ELCA, 2020).
Primary research data sources.
Consistent with an expanded Collective Intelligence Genome concept previously suggested, we considered three subsidiary questions in analysing the data: 1. Where and when to burn (establishing context) 2. What to burn and how (establishing process) 3. Who should manage the burning and why (establishing responsibility and motivation)
Findings
A brief outline of the indigenous practice of ‘cultural burning’ technology follows. Small patches of grassland, forest understory or wetland are burned when the time is right, creating a mosaic of patches that may (or may not) subsequently be connected. Patches of unburnt land may be left for a time as a temporary habitat for local fauna. Timing is influenced by observations about the current state of the flora and ambient weather conditions. Time is also important – the process is relatively slow, but this gives time for fauna, including insects to move away from the area. Time is given for regeneration and if unwanted flora appear then there may be an additional burn to remove them. Fire science researchers have noted that temperatures at the soil surface have reached 600°C in a very hot burn, for example, associated with a wildfire, effectively sterilising the soil and forming a barrier to water absorption. Temperatures may reach 100°C–250°C in a hot burn, many seeds, young and weaker perennial grasses are destroyed, and the topsoil appears charred and bare. In a cool–moderate burn, as practiced in cultural burning, soil temperatures reach 50°C–150°C. The soil 15 mm below the surface is usually not heated by more than 10°C and returns to its original temperature within 5 min. Plants that bury their seed or that have growing points below the surface have better survival after fire. Grazing animals may be attracted to new growth after a cool burn. The reduction in fuel load may inhibit the spread of wildfires. Whilst indigenous people may not have had this specific data, based on a multiplicity of observations over a long time, they were aware of the outcomes. Tacit knowledge about the process has traditionally been passed on through story-telling and direct observation over a number of fire seasons where conditions may differ.
Whilst pre-fire season understory burning is practiced by institutional fire authorities to reduce fuel load, the conventional technique involves faster burning with higher ground temperatures. Whereas western societies see deliberate fire-lighting as a risky business with regulations and approved practices to exercise control, aboriginal communities see cultural burning as an essential land management tool, with action informed by observation of the current state of the local flora and weather conditions. That had been slowly changing, with the practice being trialled in a few areas. A government review of wildfire management following massive wildfires in 2019 endorsed the practice (ECLA, 2020), but then the question was how to best share the knowledge and implement the practice more broadly.
Many researchers (e.g. Standley, 2019; Stillwell, 2010) have noted complexities in indigenous knowledge capture and diffusion as illustrated in Table 1, and there is a separate field of study associated with this activity (e.g. Cahir et al., 2016). One government approach to the collection of cultural burning knowledge was to establish an Aboriginal research enterprise (https://aiatsis.gov.au) to help undertake that task.
Where and when to burn: Establishing context
We have noted the importance of temporal considerations – identifying appropriate time windows and burn time duration, and taking a long-term view, perhaps taking a decade or more to manage a succession of cool burns that cover a large area. This may be inconsistent with institutional scheduled annual programs organised for the sole purpose of fuel load reduction. Where and when to burn is also subject to seasonal considerations, and in projects capturing related indigenous knowledge, a divergence from broad community concepts has been recorded. Whereas the tropical north has been viewed as having two seasons (wet and dry) and the cooler south four seasons (summer, autumn, winter and spring), most aboriginal communities identify with six seasons (e.g. BOM 2022). Each season has sub-ordinate events observed in the flora and fauna that signal the start and end of each season, and there are rules about how the local community should engage with nature at these times (hands off or hands on). Cultural burning is signalled as a hands-on activity in one or two time windows (e.g. McKemey et al., 2020). Here is an example from a South-Western Australian region. Birak – season of the young (BOM, 2022). First summer: December–January Mosaic burning time. ‘Birak season sees the rains ease up and the warm weather really start to take hold. The afternoons are cooled by the sea breezes that abound from the southwest. This was the fire season, a time to burn the country in mosaic patterns. An almost clockwork style of easterly winds in the morning and sea breezes in the afternoon, meant that traditionally this was the burning time of year for Nyoongar people. They would burn the country in mosaic patterns for several reasons including fuel reduction, increasing the grazing pastures for some animals, to aid in seed germination for some plants and for ease of mobility across the country.’
What and how to burn: Establishing the goal and process
One university cultural burning researcher observed ‘it was amazing to watch the men skilfully burn the landscape. Flowering was protected… there was an increase in the diversity of the understory, there was a decrease in scar height’. Comment was also made about different ways of ‘knowing’ (Standley, 2019). As knowledge about cultural burning is externalised, it is also being shared via websites (e.g. The Cultural Burning Knowledge Hub https://www.firesticks.org.au). A recent common theme that aligned the efforts of disparate community groups in responding to climate change issues was ‘caring for the land to optimise carbon sequestration practice’. This involves both increasing natural carbon sink capacity and minimising CO2 release from wildfires (e.g. Neale et al., 2019). Some indigenous national park rangers have described their modern practice as utilising two ‘toolboxes’; one drawing on traditional knowledge and the other utilising a combination of helicopters, small incendiary devices and satellite data to work in remote areas.
Who influences the burning and why: Establishing contributors and rationale
Actors, their roles and motivations for contributing to cultural burning collective intelligence.
aNotes: Fire is seen as having spiritual significance – something to be revered for both its potential destructive effects and its utility as a resource. It is combined with knowledge about the attributes of particular flora to stimulate the growth of a diverse range of plants.
bScientific modelling of the long-term impact of cultural burning has demonstrated a net carbon sequestration benefit, leading to the establishment of an indigenous social business collaboration to trade in carbon credits (see ICIN (2022)).
Maintaining trust between actors and in the underlying technology was represented in different ways. Some examples follow. One university was so impressed with the depth of knowledge of two elder rangers they were awarded honorary doctorates, which helped build trusting relationships with the broader indigenous community (Standley, 2019). Other researchers compared 1950’s aerial survey photos of a marginal desert area where cultural burning had been practiced with satellite images taken more than 50 years later after cultural burning had been discontinued for some time. This showed a degradation in the natural habitat scale and diversity with relatively large wildfire scars compared with the previous pattern of small-scale patchworks of scars (Burrows et al., 2004). The implication is that indigenous communities can be trusted to care for the land by minimising wildfire impact and leveraging particular flora attributes.
Emergent themes
One recent theme that aligned the efforts of disparate community groups in responding to climate change issues was ‘caring for the land to optimise carbon sequestration practice’. This involves both increasing natural carbon sink capacity and minimising CO2 release from wildfires and has been supported by scientific research (e.g. Neale et al., 2019). Another theme was the emergence of hybrid indigenous/technology ‘toolkits’ for use by indigenous cultural burning specialists, for example, working in remote areas using helicopters, incendiary devices and satellite imagery. A map of emergent case study themes is presented in Figure 1. Themes emerging from the cultural burning case study.
Discussion
In the following, we bring together observations from theory and from practice, introducing models to help represent interactions between matters of structure and agency. Our research question considered the use of indigenous knowledge for mutual benefit in the pursuit of SDG’s and we have considered one application – the use of fire informed by collective intelligence as an example.
The literature seems oriented towards the accumulation of knowledge from multiple sources to support complex decision-making or to establish an accessible source of truth. In our case, the relevant knowledge has been accumulated by a multiplicity of observations over thousands of years. It was personally held by small communities of individuals who shared it through a mix of story-telling and practical demonstration. Its utilisation however was time and place dependent and this is seen as a shortcoming of the original Malone et al. (2010) Genome which may have considered these attributes as particular to a specific instance. Another aspect of context – the decision environment may be seen differently by different actors and this is briefly discussed. A macro level situated collective intelligence genome model that includes time and place ‘genes’ is shown in Figure 2. A situated collective intelligence genomics model (adapted from Suran et al., 2020).
Based on prior research we note in passing that the combination of the six viewpoints represented in Figure 2: what, why, where, when and how has been used in a variety of settings. Examples include supporting creative problem solving (Vernon and Hocking, 2014) and the definition of information systems architecture (Zachman, 1987). In our case analysis, this provided a convenient framework for characterising aspects of the case. The combination has been eloquently introduced in an extract from a much quoted poem by Rudyard Kipling (The Elephants Child, circa 1920):
I keep six honest serving-men (They taught me all I knew);
Their names are What and Why and When And How and Where and Who.
I send them over land and sea, I send them east and west;
But after they have worked for me, I give them all a rest.
In Figure 2, two over-arching generic ‘genes’ (where and when) are added to the previously identified pool, and interconnections between ‘genes’ are shown. Each of these extensions to the model suggested by Suran et al. (2020) will now be discussed in the context of both collective intelligence establishment and its use.
The Suran et al. (2020) model noted that the ‘what’ gene included consideration of types of goal as individual or community. Our case and the work of others (e.g. Karsten and Illa, 2005; Ode and Ayavoo, 2020) go beyond this and consider a situated community having a ‘where’ gene that has the following attributes: • Geographical location that may be associated with specific sustainability goals (the ‘what’ gene). In our cases, cultural burning practice was adapted to local weather and flora attributes. • A particular community is situated in a political, economic, societal, technological, legislative and ecological (PESTLE) space (e.g. Zalengera et al., 2014) that influences genes in relation to resource access and rules. In our cases, cultural burning may be allowed in conjunction with government authorities in particular places, for example, tribal land. • Suran et al. (2020) identified a ‘decide’ attribute as part of the ‘how’ gene. We have previously observed that a particular goal may be associated with a particular decision environment: ordered or unordered, that may be seen differently by different actors. For example, the existence of wildfires may seem as chaotic to some, but indigenous communities have witnessed this over multiple generations and see it as normal.
We have proposed the addition of a ‘time’ gene as an element of context. The Ancient Greeks had described two views of time. Chronos, a linear process founded on astronomical observations that may be represented as divisible units used to synchronise activities, and Kairos, a place-dependent view of timely action founded on socio-ecological observations. We adopt the latter view which can accommodate elements of Chronos, and it has been utilised in innovation studies (e.g. Beckett and O'Loughlin, 2016). Our addition of a ‘time’ gene has the following attributes: • Events in time, for example, milestone events or turning points that may be linked to the ‘why’ gene. In our case, deciding when to burn. • Lifecycle event duration and re-occurrence, for example, in nature or in the utility of an innovation that may be linked to the ‘how’ gene. In our case, indigenous Australians view annual cycles of six seasons. • Time windows, opportunities for action, for example, in response to an emergency (Mendonça et al., 2000). In our case, combined observations about weather and flora conditions indicated appropriate time windows. • Timely action, considering both the past and the future – when to act and when to wait (e.g. Phillp and Martin, 2009; Torbert and Taylor, 2008). In our case, deciding where not to burn was just as important as deciding where to burn.
Interactions between ‘genes’ are noted in the forgoing. Suran et al. (2020) had identified interactions between ‘who’ and ‘how in their model. ‘Who’ provided input to 'how’ in accumulating collective intelligence, and this is observed in the establishment of both indigenous community stories and in establishing the scientific rationale associated with them. ‘How’ provided an input to ‘who’, in our cases, identifying actors knowledgeable about aspects of cultural burning in a particular place. We identified other interactions, shown in Figure 2: • Who has the prior knowledge to support goal achievement and who has the absorptive capacity to appreciate the utility of new knowledge related to goal achievement. In our cases, indigenous community actors and selected ‘apprentices’. • Who influences the prioritisation of particular actor engagement and why. In our cases, government actors may ‘certify’ both indigenous and non-indigenous actors to be involved in cultural burning activities. • Which goals are assigned a higher priority and why. In our cases, government priorities for cultural burning in state or national parks were associated with keeping access roads clear, whilst indigenous people may favour the protection of sacred sites. • How does the goal influence process selection. In some of our case examples, the goal of treating a remote area required the use of helicopters and satellite imagery as tools. • How do outcomes from operation of the process influence motivation for its future use, and what motivations drive the selection of a particular process. This may be characterised as demonstrating value-in-impact, but one potential problem is that this assessment may take some time to be realised (e.g. if there are no subsequent wildfires in an area treated using cultural burning).
These observations about interactions between views prompted us to consider a future study of an interaction matrix to show interactions more comprehensively, an approach that has been utilised in sustainability studies (e.g. Vázquez et al., 2015) and in collective intelligence studies (e.g. Chang et al., 2015). The interactions seem to be about decision-making, and this view may be a topic for future research.
Salminen and Harmaakorpi (2012) had suggested that human collective intelligence processes may be considered at three levels: a macro-level (we have represented in Figure 2), a micro-level where agency behavioural elements and trust may influence engagement and an emergence level where process adaptivity, interactions and rules emerging from macro-level processes are reinforced. Consistent with theoretical observations, our case study indicated that emergent-level practice is embedded in a combination of context-dependent rules and access to supplementary knowledge as a resource. Some observers of cultural burning in action have been perplexed by what seems to be a series of random decisions (e.g. Verran, 2002). There seemed to be more than one logic in play, more than one goal to be achieved (e.g. reducing fuel load and protecting particular plant species). If a field observation is interpreted and validated, then a specific action takes place leading to an outcome that has some impact. A combination of both agency and structure is required. The processes of interpretation (sense-making) and validation (legitimation) condition what action is appropriate. Reflection on that outcome (learning) may lead to rule reinforcement or adaptation. This resonates with the concept of ‘situated reflective practice’ observed by Malthouse et al. (2014) in working with life-long learning educators where an interplay between agency, structure and time-place context was noted. This may be a topic for further research. Another outcome may be an enhanced knowledge resource including an appreciation of timing and enhanced absorptive capacity of the actors involved. The temporal aspect of knowledge has been highlighted in the development of intelligent systems (e.g. Maniadakis and Trahanias, 2011), and may be a topic for further research. Figure 3 illustrates firstly, the conditional decision processes at work, bringing together indigenous knowledge drawing on rules and secondly, some potentially relevant theories that may be associated with the observed practice that may also indicate themes for future research. Collective intelligence rules and knowledge informing action: An indigenous knowledge case example.
The lower part of Figure 3 introduces a learning component consistent with the collective intelligence intrinsic motivation of seeking and sharing knowledge (Table 4). It reflects the importance Australian indigenous people place in learning ‘on-country’ as an adjunct to story-telling in building the absorptive capacity of novice active contributors.
We suggest that just as we all use pattern recognition to stimulate action, so do cultural burning actors. We check the weather to inform how we get dressed. We pick out a friend in a crowd and organise a meeting. Indigenous actors’ observations on the behaviour of fauna and flora establish the start of a new season where cultural burning may be appropriate. Observations about flora and weather conditions at a particular time and place inform the decision to burn (or not). There are some parallels with regional fire authority practice of only permitting prescribed burning in particular areas prior to the ‘fire season’, and there are rules reflecting requisite prior experience. But there are also differences in implementation arrangements that can be a source of tension.
As noted earlier, experience-based stories are used as a resource in the utilisation of case-based reasoning tools and pattern language tools that inform context-specific action. Storytelling has been used for a long time as a tool for introducing individuals and groups to rules that support social coherence and minimise risk. Examples are European fairy-tales and religious texts. Reflection on the stories with the aid of a knowledgeable interpreter is needed to surface and reinforce these rules. We suggest there are some parallels between the structure of stories and our extended collective intelligence genome individual ‘genes’. By way of example, one of the authors of this paper has used a computer-based application (Storymill) to help construct book-length articles for personal use. This tool invites the author to organise the book into chapters, and for each chapter research scenes (where and when), characters and their roles (who was involved and why were they important in the story) and key events (what took place and how this influenced the story). Put another way, information is organised into small chunks that may be configured for a particular purpose. We suspect that this configuration activity – bringing together particular pattern sets – is what happens within the AND conditional functions shown in Figure 3, with the rules being associated with where and when, and the resources being associated with what and how, supporting the rationale (why) for a particular action by a particular agent (who). Two examples from the literature follow. Barnhardt and Kawagley (2005) noted how the detailed observations of indigenous Alaskans about patterns in animal behaviour led to the evolution of sophisticated hunting techniques. Roos (2013) described how a detailed knowledge of the behaviour of eels during times of high rainfall led to the establishment of a form of indigenous aquaculture. Akasaka et al. (2020) used a pattern language approach to draw on the collective intelligence of 51 experienced ’living lab’ practitioners in Denmark and Japan to identify 30 patterns representing experience in (a) co-creation with users, (b) co-creation process management and (c) team building for co-creation.
The points we make here are that there is an underlying compatibility between elements of storytelling as a knowledge diffusion mechanism and the ‘genes’ of the situated collective intelligence genome presented in Figure 2, and that they may be utilised in sets to show patterns. The use of pattern language to help capture and utilise indigenous knowledge for mutual benefit may be a topic for further research.
Concluding remarks
Our research question was ‘how might indigenous knowledge that could support the pursuit of SDGs be represented from a mutually beneficial collective intelligence perspective’? The short answer is through a combination of structure and agency drawing on the accumulated knowledge as rules and knowledge as a resource. Mutual benefit may be derived by combining indigenous experiential knowledge with scientific research knowledge given that ethical practices are followed. Such practices have been described as ‘two-eyed seeing’ which has emerged as a specific field of research (e.g. Smith, 1999).
Indigenous people have accumulated knowledge about what works (or does not work) and why over periods of hundreds or thousands of years. Who retains this knowledge, how it is utilised and shared, where and when it is used are associated with continuous observation of the natural world the community is embedded in and on the short and long term needs of that community. ‘Caring for country’ and learning ‘on country’ are culturally embedded practices.
Dramatic increases in the global population through the 20th century has impacted on nature and societal needs, and changes nature (e.g. in relation to climate change) and societal needs (e.g. in protecting disadvantaged communities) have led to a United Nations pronouncement of 17 sustainable development goals. Goal 17 is about revitalising global partnerships for sustainable development and ensuring no one is left behind, and we see that collaboration between indigenous and non-indigenous communities as part of that.
We drew Australian case studies of cultural burning practice supporting the achievement of SDG 13 – ‘take action to combat climate change and its impacts’ – to gain insights into matters of agency and structure in the utilisation of collective intelligence. The literature review of indigenous knowledge capture and utilisation revealed a multiplicity of factors to consider (Table 1). Two factors motivated ongoing learning via a number of different mechanisms. Firstly, there were concerns that this indigenous knowledge may be lost, and secondly indigenous knowledge may be combined with scientific knowledge and the use of supplementary technology tools to more efficiently deliver the desired outcomes. The Australian indigenous community’s historical motivation has been caring for the land to support long-term food source sustainability, combining cultural burning with botanical management. Recent research has identified another benefit from these practices – a net carbon sequestration benefit where trading in carbon credits also provides a financial motive for the indigenous community to continue the practice. Again, there were multiple factors to be considered as summarised in Figure 1. Working with fire in open landscapes can be seen as a risky business, and trust in the utility of the practice and the actors involved had to be developed over time. Although broad community interest had been slowly accumulating, a major incentive was provided from reviews of the impact of massive wildfires in 2019, where evidence of reduced impact in areas previously subjected to cultural burning practice was accumulated.
Collective intelligence researchers have suggested such action supports an accumulation of historical knowledge with the goal of (a) establishing rules and an enhanced knowledge base, (b) in support of complex decision-making or (c) in evaluating future possibilities/options. Using that knowledge in conjunction with emergent technology (e.g. satellite mapping) provides an opportunity for on-going learning. As a better understanding of the science underpinning indigenous knowledge evolves, the combination may identify future possibilities. A counter-intuitive example mentioned earlier is that the practice of cultural burning offers a net carbon sequestration benefit, which may be verified with the aid of satellite technology.
Statement of significance
The notion of a ‘genome’ structural model associated with collective intelligence projects has previously been applied to a variety of situations, and we have extended the original concept identifying who, what, why and how ‘genes’ to include two aspects of context – time and place ‘genes’. We have also extended observations about interactions between the genes in a situated collective intelligence gene model (Figure 2). We suggest that decision-making is not just associated with the ‘how’ gene as represented in the original concept, but is associated with connections between the ‘genes’ that may be viewed as a complex matrix of decisions. Drawing on experience in another field, case material was examined from three perspectives: what knowledge was applied and when, who was involved and why (e.g. see Table 4) and where and how was the knowledge applied (e.g. in bushland, savannah or wetland locations). This provided a rich representation of the knowledge underpinning the practice of cultural burning. Indigenous knowledge application has been represented as rules and a knowledge base, together supporting if-then decision-making in a complex environment (Figure 3). The application of these rules and utilising knowledge as a resource also has an experiential learning dimension based on what works and what doesn’t. Matters of context – where and when this knowledge can be utilised and the nature of timely decisions to be taken in parallel are seen to be elements of our extended collective intelligence genome model (Figure 2).
There are also matters of agency to consider. Whilst the case study practice of cultural burning had been utilised by indigenous communities for thousands of years, it was discontinued in many parts of Australia, as it was seen to be risky. It has only been recently endorsed by institutional actors as a valid practice following a better understanding of the underlying science. This indicates there may be cultural barriers and absorptive capacity issues that impact the utilisation of indigenous knowledge, and it may take some time to make the transition. Indigenous knowledge has been captured and shared through story-telling and experiential learning to show the utility of the underlying messages. We suggest there are logical connections between the components of our extended genome and the development of stories: the scene (where and when), the actors (who is involved and why) and the associated events (what happened and how). In addition, we note the presence of patterns within and between stories, for example, as utilised in case-based reasoning tools, and briefly discuss the potential utility of pattern language in characterising indigenous knowledge.
Opportunities for further research
In this study of collective intelligence associated with indigenous knowledge, we touched on a number of theoretical viewpoints that may help in the representation and understanding of that knowledge. In that context, some suggested opportunities for further research are • The relevance of Copeland and Moore’s (2018) four dimensions of trust (legitimacy, authenticity, synergy and commons) in supporting ‘two-eyed seeing’. • Further investigating interactions between collective intelligence ‘genes’ (who, why, what, where, when, how). • Considering the relevance of Malthouse et al. (2014) ‘situated reflective practice’ to better understand the dynamics of indigenous collective intelligence. • Better understanding temporal factors associated with collective intelligence, including Maniadakis and Trahanias’s (2011) temporal cognition theory. • The utility of Giddens (1984, 1985) structuration theory in characterising the rules component of collective intelligence. • The role of absorptive capacity in facilitating indigenous – non-indigenous community knowledge exchange. • The use of pattern language tools to help capture and utilise indigenous knowledge for mutual benefit.
Footnotes
Acknowledgements
We thank Troy MacDonald, the then Board Chair of the Gunaikurnai Land and Water Corporation for his support in identifying sources of Australian indigenous research into the practice of cultural burning. Secondly, we thank the reviewers of our original submission who suggested ways the models presented might be refined and who reminded us of the importance of considering matters of agency when working with indigenous collective intelligence.
Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
