Chapter 5. The uses of knowledge
p. 69-80
Texte intégral
“If […] the problem of society is mainly one of rapid adaptation to changes in the particular circumstances of time and place [then] the ultimate decisions must be left to the people who are familiar with these circumstances, who know directly of the relevant changes and of the resources immediately available to meet them. We cannot expect that this problem will be solved by first communicating all this knowledge to a central board which, after integrating all knowledge, issues its orders. We must solve [the problem] by some form of decentralisation […].
But the ‘[people] on the spot’ cannot decide solely on the basis of [their] limited but intimate knowledge of the facts of [their] immediate surroundings. There still remains the problem of communicating to [them] such further information as [they need] to fit [their] decisions into the whole pattern of changes of the larger […] system. How much knowledge [do they] need to do so successfully? Which of the events which happen beyond the horizon of [their] immediate knowledge are of relevance to [their] immediate decision, and how much of them need [they] know?”
Friedrich Hayek (1945)
1Like many academics involved in global health, I am frequently called upon to make judgements on research papers (Abimbola, 2018). To assess their value, I rely on their declared aim, which is usually stated at the end of the “Introduction” or “Background” section. The declared aim can reveal for whom we imagine we write (i.e. gaze or audience) and the standpoint from which we write (i.e. pose or positionality). But much too often, it reads like an afterthought, as though it was written in pretence or in tenuous hope, like an item thrown into deep space in the hope that it might hit a target called “equity” or “improving the lives of people in need”. The declared aim of research papers in global health often betrays our lack of a collective sense, or theory, as to how knowledge leads to equity – a theory with which we may align how we value, produce and use knowledge. Such a sense of how knowledge does its work exists in the basic and biomedical sciences, but not in global health – or in a more prosaic rendering – not in our efforts to achieve health equity.
2Much of the edifice of academic global health is built on the assumption that the goal of inquiry is to identify universal truths: to add to a central pool of knowledge (Rutter et al., 2017; Greenhalgh, 2020; Ogilvie et al., 2020), e.g. on the biology of a virus, or the quest for a vaccine. The propensity to value the universal in global health – often apparent in the declared aim of research papers – stems partly from its biomedical and colonial origins (Affun-Adegbulu & Adegbulu, 2020; Biehl & Petryna, 2013, 2014; Richardson, 2020). In biomedicine, you search for universal truths insofar as biology, chemistry and physics are essentially the same everywhere. In colonial relations, you look from a distance, and you are unable to see or value what is unique to a place. You see whatever you deem universal or value what you see insofar as it aligns with your foreign, external worldview. You conflate what it means to be “rigorous” with a thought-stopping cliché that demands universal truths and external validity, a randomised controlled trial and approximations of it, or a default to uniform and standardised methods, metrics, indicators and measures around the world. But truth, in such a centralised or universal sense, is typically not the goal in the quest for equity within and across social spaces and systems (Rutter et al., 2017; Greenhalgh, 2020; Ogilvie et al., 2020). What often matters much more is finding contingent, circumstantial or particular truths about what it takes to achieve equity at a particular place and time, and subsequently, finding regularities; finding the universal in the particular. After all is said and done, eventually, ultimately, you can go looking for regularities across places and times. But only after finding the specific, and then abstracting across multiple specifics.
3If all you read are research papers in elite academic journals, the impression you get from the literature is very likely that equity in global health comes about primarily through “surgical” change brought about through discrete, episodic, one-off events or efforts. That is, macro-innovations or (often externally introduced) measurable, tangible, time-bound interventions, which function independently of one another and of context, and whose effects can be seen from a distance (Biehl & Petryna, 2013, 2014; Richardson, 2020; Storeng & Béhague, 2014; Weick, 1984). You may also get the impression that large or multi-site studies are inherently more valuable than small or single-site studies; or that there is virtue in adopting methods and measures whose primary purpose is to simplify complex realities. The literature is a misleading tip of an iceberg. What remains hidden beneath the surface are the day-to-day uses of knowledge by actors working within systems at different scales of organisation to bring about change: that is, “organic” change. The literature thus reads like a conversation to which the primary participants, those working for change day to day, were not invited.
4The academic global health literature marginalises many conversations that should be primary. Conversations on the role that social learning plays in how knowledge is used to achieve health equity (Sheikh et al., 2020; Ellerman, 2001; Schon, 1994). Conversations within and between countries on the organic processes that underpin equity in health and its determinants, which, even when local, would often have international resonance. It is time to look again at our assumptions of the primary purpose of our literature, and to consider what a reordering of those assumptions might mean for how this literature is put together. The status quo reflects a pervasive misordering of value (Richardson, 2020). We need to reorder our working assumptions so that they begin with a (rebuttable) presumption that there are primary uses, users and producers of knowledge that our literature must serve first and foremost. The construction of such a different order for our literature – for how (or how not) to use knowledge for equity – could be based on the principle of subsidiarity.
——
5“Subsidiarity” comes from subsidium, a Latin term that originated in the Roman military. It was used to describe the third line of soldiers, “who would only join in a battle if the powers of the two front lines were insufficient” (Gosepath, 2005). The military connotation of subsidium has faded over time. The term now refers to help, assistance, support or aid – the kind that one would not need under ideal conditions, or the kind only offered on request (Gosepath, 2005; Cahill, 2017; Höffe, 1996). In its modern usage in social and political philosophy, “subsidiarity” is used in relation to an ethical principle: the principle of subsidiarity. This principle holds that decisions about efforts to help others and to attain the common good (for example, by using knowledge to achieve equity in global health or within a social space or system) should, by default, take place at the smallest or most proximate level/scale of organisation possible, and only when necessary at a larger or more distant level/scale of organisation (Gosepath, 2005; Cahill, 2017; Höffe, 1996; Chaplin, 1997).
6Evoked as a safeguard against (over)centralisation, subsidiarity is a priority principle for relations between governing entities. Subsidiarity means (until proven otherwise) that smaller/proximate units take precedence over larger/distant units – individual nations over supranational entities to which they belong, subnational over national governments, communities over governments, and families over communities. The principle of subsidiarity predates its modern, explicit usage. For example, subsidiarity was implicit in the formation and relations within and between loose confederations of historical Greek and Yorùbá city states (Usman & Falola, 2019; Adelusi-Adeluyi & Bigon, 2008; Rufus Davis, 2021; Roy, 2007). In global health, there is a multiplicity of proximate–distant (or primary–subsidiary) relations (between social, technical and political entities within and beyond jurisdictional boundaries). Subsidiarity provides a practical basis for governing aspects of these relations; especially those that pertain to the uses of knowledge.
7The case for adopting the principle of subsidiarity to reorder our assumptions about the uses of knowledge can be made on two inseparable grounds: on the practical (e.g. effectiveness and efficiency) and moral (e.g. equity and justice) benefits of acting close to the ground, close to the issues, and close to the people. Among the practical benefits are that this principle allows proximate actors with optimal knowledge of the problems to take initiative, to experience the consequences of their actions, to revise their theories, and to inform subsequent iterations of action based on local information, feedback and observed regularities (Gosepath, 2005; Cahill, 2017; Höffe, 1996; Chaplin, 1997; Drew & Grant, 2017; Abimbola et al., 2019). By promoting proximate initiative and action at or by diverse smaller scales/entities, this principle also promotes plurality and multiplicity of forms, which in turn increases opportunities for learning (within and across such small units) and for resilience to shocks (given network effects across small units) (Höffe, 1996; Drew & Grant, 2017; Abimbola et al., 2019).
8The moral benefits of the principle of subsidiarity mirror its practical benefits, as moral proximity tends to follow physical proximity. Moral proximity implies accountability to oneself and to one’s community, with a high stake (“skin in the game”) in ensuring the success of collective efforts or initiatives. Putting one’s own proximate (often tacit and informal) knowledge to use can inspire one’s confidence in its value. And with proximate action and use of knowledge, actors see how their (often collective) efforts help to promote the common good. This experience can then help them achieve the deeply human and existential need to be useful, to avoid social alienation and dependency, to exercise agency, and to gain the sense of dignity and meaning that comes with shaping one’s own destiny (Drew & Grant, 2017). Indeed, the moral necessity of the principle of subsidiarity is such that assigning to a distant or subsidiary entity what a proximate or primary entity can do has been described as “a grave evil” and “an injustice” (Pope Pius XI, 1931).
9On the road to equity, knowledge is primarily produced or used by proximate actors to generate small wins – continuous day-to-day “organic” micro-innovations (Weick, 1984; Ellerman, 2001). On the other hand, subsidiary production or use of knowledge is often about large wins, typically episodic, “surgical” or major innovations, designed or enacted at a distance. Proximity focuses attention on small wins. Large wins are really a cumulation of these small wins, of minor changes that result from ongoing learning. Small wins require less coordination to implement, and they are more structurally resilient compared to single large wins. Each small step inspires confidence in the possibility of success in the next step. These steps can be assembled in endlessly varied combinations from place to place and from time to time. But what may appear to a distant subsidiary actor as the “next logical step” or “next solvable problem” typically diverges significantly from that of a proximate primary actor (Weick, 1984; Ellerman, 2001).
10The Irish legal scholar Maria Cahill has distilled the principle of subsidiarity into four precepts (Cahill, 2017). These four precepts may be applied to the uses of knowledge along the following lines (Gosepath, 2005; Cahill, 2017; Höffe, 1996; Chaplin, 1997). (1) Primary units must have the freedom, opportunity and responsibility to produce, use and rely on their own knowledge, and to indicate when they need the help of subsidiary units.
(2) Subsidiary units must only provide help when requested, respecting the presumptive authority of primary units in relation to knowledge, unless primary units are evidently so weak that they are unable to seek help.
(3) Subsidiary units must direct requested or offered help towards the goals of primary units and not their own goals. (4) There must exist built-in structures to govern the engagement of subsidiary units with primary units, e.g. to decide when to modify or (dis)continue help or what to do with knowledge that is produced, so that subsidiary units do not undermine primary units and primary units do not become overly dependent on subsidiary units.
11How then may we distinguish primary from subsidiary uses (or units of users or producers) of knowledge? The principle suggests that academic research is subsidiary to knowledge from the arena of practice; policy designers are subsidiary to implementers; what is episodic is subsidiary to what is day to day; and distant decision-making is subsidiary to proximate decision-making. These distinctions suggest four uses (or units) of knowledge (Figure 8) (Dekker, 2020). Two uses/units are primary: emancipators (people using knowledge to transform the structural determinants of their own health) and plumbers (implementers using knowledge for change). Two uses/units are subsidiary: engineers (policy designers using knowledge from primary units) and professors (knowledge-mongers). Each use/unit may exist in an individual, group or organisation. Quite like fractals, the units of four form a knowledge ecosystem that is present at different scales of organisation (Figure 9).
12“Emancipators” are activists – individual activists, teams of campaigners, women’s groups, civil society organisations, etc. They use knowledge in their efforts to construct a new reality for themselves, to demand and create ideal conditions (economic, social, political, gender) for themselves, and to alter the social structures and rules that disadvantage them (Galtung, 1969), regardless of whether those structures and rules are made locally or at a distance. In the process, as they work to transform social systems and realities, they also produce knowledge.
Figure 8 | Units of uses/users of knowledge in global health.
Activist/Constructive Attitude | Accepting/Corrective Attitude | |
Proximate/Local/Primary Position | “Emancipator” | “Plumber” |
Distant/Foreign/Subsidiary Position | “Engineer” | “Professor” |
The distinctions drawn here are partly inspired by Dekker (2020) and Abimbola (2019).
13On the other hand, broadly accepting of the status quo, “plumbers” work within existing systems. They use and produce knowledge day to day, bringing about incremental change through service delivery as individual bureaucrats, teams of service providers, NGOs, or ministries of health. As primary units, plumbers and emancipators are together responsible for the vast bulk of change in global health (Weick, 1984; Sheikh et al., 2020; Ellerman, 2001; Schon, 1994; Dekker, 2020; Galtung, 1969; Gumbonzvanda et al., 2021; Levitt et al., 2013; Hendel & Spiegel, 2014; Askew et al., 2020). They make progress in small wins. They learn from these small wins, and from small losses too. They learn by doing.
Figure 9 | Concentric rings of knowledge uses/users in global health.

This figure shows (a) examples of the scales of organisation at which knowledge is used for equity (the five concentric rings); (b) the spaces in which the knowledge is used (the small circles that line the concentric rings); and (c) the four uses/users or units of knowledge, which are represented as four labelled circles within each small circle. The size of the primary uses/users or units (emancipators and plumbers) is larger closer to the community “ring”, while the size of subsidiary uses/users or units (engineers and professors) is larger closer to the global “ring”.
14“Engineers” design policies, programmes and systems, or they create the overarching framework within which plumbers perform these and their own roles (Dekker, 2020; Erasmus, 2014). “Emancipators” seek to alter the designs of engineers. Engineers have power or align with power. They are considered “activists” insofar as they can use their power to construct emancipatory realities, especially when emancipators push engineers to transform unjust social systems and realities. Engineers are visible. They attract the attention of “professors”. Too much of our literature is aimed at engineers, as if change emanates primarily from the intercourse between the political and academic elite – as if the literature is just an instrument of power. The knowledge needs of (and the knowledge produced or held by) emancipators and plumbers go unacknowledged, as if they do not exist. Yet the vast bulk of the literature should be written for and by emancipators and plumbers. If this claim appears farfetched, it is because the status quo is not fit for the purpose.
15“Professors” can be individual academics, journalists or reporters, town criers or village announcers, research or data teams or groups, think tanks, academic institutes, or universities. Or the research or data team within an organisation of engineers, emancipators or plumbers. Professors share knowledge through publications and teaching and activities that help to connect a system, any system, “to more of itself” (Wheatley & Kellner-Rogers, 1998; Wheatley, 2001; Wheatley & Frieze, 2015). Disconnection is the source of many inequities. To connect a system “to more of itself” is to make unavoidably known – elsewhere or by others – what was already known or has been experienced by a part of a system, or a set of actors or knowledge users. It is to connect one voice to another voice, and many voices to many ears. Professors also repackage and reframe existing knowledge, and sometimes make new knowledge. To contribute to equity, professors might seek to undo assumptions that hold it back or develop analytical tools to better define and solve problems (Whyle & Olivier, 2020). Some, in an ivory tower sense, may be uninterested in equity, curating seemingly useless knowledge, some of which may nonetheless become useful for equity. But perhaps the most direct – and primary – way in which professors can use knowledge for equity is by helping to connect a system to more of itself; by moving knowledge within and across time and place.
16The academic global health literature has predominantly been a space for professors. It is time for that to change. What engages the attention of professors is not necessarily what advances equity in global health. Professors are party to equity when they are part of the learning processes of and between plumbers and emancipators (with or without engineers), purposefully helping to connect units to one another. They may be professor–emancipators (scholar–activists) (Gumbonzvanda et al., 2020; Eschle & Maiguashca, 2006; Musolino et al., 2020; Campbell & Burgess, 2012) or professor–plumbers (scholar–implementers) (Irimu et al., 2018; Ropa et al., 2019; Young et al., 2018). They may work directly with engineers to make sure that their designs are based on knowledge produced or held by plumbers and emancipators (Barasa et al., 2020; Montilla et al., 2020; McAteer et al., 2019; Uzochukwu et al., 2016). Even then, engineers need to aggregate a plurality of interests and knowledge (Belghiti Alaoui et al., 2020; Bonvin & Laruffa, 2018; Salais, 2008). Professors may help to connect different interests and knowledge by facilitating platforms for deliberation among groups of emancipators or groups of plumbers, or between emancipators and plumbers (Bryant, 2002).
17Equity is political. Engineers and emancipators are inevitably political. Even plumbers will sometimes have to be emancipators (Schaaf et al., 2020; Abers, 2019; O’Sullivan, 2015; Brown & Fee, 2006). Professors will often need to be political. Uses and users of knowledge can also be fluid. An emancipator for one cause can be a plumber or an engineer for another. An engineer for one design may be a plumber or an emancipator for another. But the use of knowledge by engineers for “big wins” is so visible that it is far too often the focus of professors’ attention (Biehl & Petryna, 2013, 2014; Richardson, 2020; Storeng & Béhague, 2014; Weick, 1984). The use of knowledge by plumbers gets a lot of professors’ attention too, but nowhere near enough. Emancipators get even less attention. The primary uses of knowledge (by plumbers and emancipators) often go unseen, in part because despite small wins and the learning that results from them, day by day nothing appears to change (Weick, 1984). Change is more apparent in hindsight. Much more visible is the episodic work of professors (aimed at engineers) and of engineers (reported by professors).
——
18There are countless spaces within which knowledge may be used to connect a system to “more of itself”. Such spaces exist at different scales of organisation – communities, districts, subnational or national entities, the globe (Figure 9). Inside each space, actors produce and use knowledge as primary units, sometimes with the help of subsidiary units. Knowledge from small wins may get shared within the space in which it is produced, or shared between spaces at the same scale of organisation (e.g. between districts or countries) or at different scales of organisation (e.g. between a community and a district, or a country and a global entity). The principle of subsidiarity suggests that the flow of knowledge should circulate within spaces, serving proximate spaces and primary units before it is put to subsidiary uses. Maria Cahill’s four precepts of the principle of subsidiarity (Cahill, 2017) suggest how we might make sense of and limit injustices in the uses of knowledge.
19First, injustice occurs when subsidiary units fail to default to primary units as primary producers and holders of knowledge. Or when subsidiary units do not allow primary units to make the most of the knowledge that they produce and hold. It manifests when subsidiary units default to a deficit-based approach to knowledge rather than asset-based or strengths-based approaches to knowledge (e.g. reckoning with existing knowledge, efforts, and small wins) (Morgan, 2014). It manifests in failure to minimise the need for subsidiary units, by not investing in platforms that allow primary units to learn for themselves, to circulate learning from small wins, and to connect to more of themselves (e.g. through routine data, deliberation or after-action review) (Ghaffar et al., 2020). To limit such injustices, it may be necessary to mandate that any help provided by subsidiary units should be provided in a way that does not take away from (but instead reinforces) the capacity and opportunity of primary units to learn for and by themselves.
20Second, injustice occurs when subsidiary units presume to have the right to help primary units to produce or use knowledge when there is no explicit request to do so and no evident lack of capacity to make such a request. The effect may be to usurp the ability of primary units to learn for themselves and develop their own learning structures as they iteratively produce and use knowledge to generate small wins. Or it may manifest in the failure to see that the only help worth offering without explicit request is to offer platforms that facilitate the flow of existing knowledge, which then connect a system, first to more of itself, then its neighbours and similar entities elsewhere (Bailie et al., 2018). To limit these injustices, perhaps ethics approval for any help provided by subsidiary units should require evidence of a request from primary units (e.g. through a representative process), or evidence of an inability to request help, or that the help being offered is only subsidiary.
21Third, injustice occurs when subsidiary units prioritise or privilege their own perspectives, concerns or needs (for example, what is “innovative” or “new knowledge” to the foreign gaze) over helping to connect a system to more of itself. This may manifest in a tendency to focus energy on subsidiary activities even when primary units request help: for example, acting as if publication is a primary goal of such an effort, even though the knowledge in publications pales in comparison to whatever knowledge is produced while assisting plumbers and emancipators, or even engineers. Perhaps academic journals, institutions and funders should insist that knowledge ought to have achieved its primary work in the world before it is allowed to be published or used for other subsidiary purposes. This way, as much about “new findings”, research papers would be about the lessons learned during the primary work – that is, while being used to connect a system to more of itself.
22Fourth, injustice occurs when there are no effective structures to govern the terms of engagement between subsidiary and primary units to avoid subsidiary units undermining primary units or primary units becoming overly dependent on subsidiary units. This governance could involve, for example, a committee to review, modify or discontinue engagement, to ensure that subsidiary units do not unilaterally disengage, and to guide what to do with any knowledge generated as part of the engagement, and what gets published, how and where. The committee might recommend that knowledge intended for primary uses should be presented concretely, while knowledge intended for subsidiary uses should be presented in an abstract way that allows insights to travel (Abimbola, 2018). The committee might be responsible for approving publications, their authorship and intended audience. It may insist on local outlets and forms that are readily accessible to the relevant primary units. It may insist on serving primary units and uses first and foremost.
23The academic global health literature needs to develop more just and inclusive ways of reckoning with knowledge of all kinds – especially from and for plumbers and emancipators. The literature must be more open to different ways of sharing knowledge. It must seek to approximate available knowledge. Ongoing concerns about lack of representativeness of authorship must also be understood in terms of lack of representativeness of the knowledge available in the literature, or lack of representativeness in whom or whose purpose the literature serves primarily.The literature needs to begin from a set of premises that defaults and defers to the practice-based knowledge produced and held by plumbers and emancipators and at proximate scales of organisation. Professors will need to invest as much time, energy and resources in plumbing and emancipation, in helping to better connect systems and in making routine data systems strong, as they do in initiating research. With such strong connections and data systems much research will become redundant.
24The assumptions that underpin our system of prestige and value in academic global health are essentially extractive, and thus colonial. These assumptions make the literature read like subsidiary actors speaking to one another in echoes about subsidiary actions. They lead us down paths of injustice. In placing the value of a study on publication, and considering that what is not published is not known, we have been unjust. In choosing to focus on the episodic at the expense of the day to day, we have been unjust. In failing to prioritise what is primary over what is subsidiary in our uses of knowledge, we have been unjust. In failing to see, share and publicise small wins, we have been unjust. In reinforcing the notion that external validity and standardised measures come before local use and local relevance, we have been unjust. In thinking that our primary role is “new” knowledge rather than helping to connect a system, any system, to more of itself, we have been unjust.
25This is not to pitch one use of knowledge against another. The principle of subsidiarity is not so much about what should be done or not done as it is about what should be prioritised. It is a call to rethink our relative focus, emphasis, priority, gaze. It is a call to default to the local gaze; to take the imaginative leap that allows a foreign (or subsidiary) actor to assume, speak or write from a local (or primary) pose. It is a call to reorder the hierarchy of rigour and value that shapes our assumptions about knowledge; to recognise where the vast bulk of knowledge used to advance equity in global health is to be found. It speaks to an intellectual deficit in academic global health (Abimbola & Pai, 2020). This deficit overlaps with the problem of colonisation, but not completely. The principle of subsidiarity shows us how we default to injustice – it also points to an alternative vision of academic global health, and its literature.

Le texte seul est utilisable sous licence Creative Commons - Attribution - Pas d'Utilisation Commerciale - Pas de Modification 4.0 International - CC BY-NC-ND 4.0. Les autres éléments (illustrations, fichiers annexes importés) sont « Tous droits réservés », sauf mention contraire.
La recherche interventionnelle en santé des populations
Concepts, méthodes, applications
François Alla, Linda Cambon et Valéry Ridde
2024
Population health intervention research
Concepts, methods, applications
François Alla, Linda Cambon et Valéry Ridde
2023
Investigación de intervenciones en salud poblacional
Conceptos, métodos y aplicaciones
François Alla, Linda Cambon et Valéry Ridde
2024