Science Africa Doesn’t Know It Has

Imagine an African scientist working on a problem in a university laboratory. Somewhere else on the continent, another researcher is working on a related problem. A third has the technical expertise needed to help turn the first researcher’s findings into something useful. Perhaps an industrial company has been looking for precisely the capability that the university laboratory is developing.

The people, knowledge and resources needed to make progress may already exist. But the researchers do not know one another. The company does not know where to look. The institutions responsible for funding science may not have a complete picture of the work taking place within their own country.

The problem, in this case, is not that Africa has no scientists, or that nobody is doing worthwhile research. It is that the scientific possibilities already present on the continent may be more difficult to see, connect and develop than they ought to be.

This is a different way of thinking about Africa’s scientific predicament. We often ask how much African countries spend on research, how many scientists they train, how many laboratories they build and how many technologies they import. These are important questions. But equally important, perhaps more so, is this one: how well does Africa know the scientific capabilities it already possesses?

Pay attention to a new development in the way science itself is being studied. Researchers are beginning to use large-scale data to investigate whether the patterns of scientific activity can reveal which research fields are approaching important discoveries. If such methods can help identify promising research before its importance becomes obvious, could they also help African countries discover scientific opportunities that are currently overlooked?^[1]

This goes beyond how governments should distribute research grants. It is a question about whether African science can become more visible to itself and whether that visibility could become part of the infrastructure through which scientific discoveries are made.

The problem of picking scientific winners

Every government that funds scientific research is confronted with the same set of questions. Should it concentrate resources on problems it has decided are especially important, or should it distribute funding broadly and allow researchers to determine which questions deserve attention?

The first approach has an obvious appeal. A government can identify pressing national needs such as disease, food security, energy, environmental degradation and industrial development, and direct research funding towards them. The country can organise its scientific resources around specific goals rather than simply hoping that useful discoveries emerge.

But scientific knowledge does not always advance according to a government’s timetable. A technology may be attractive before the underlying science is sufficiently understood. A research programme may even be organised around a mistaken assumption. And the decision to prioritise one approach can divert resources from alternatives that would eventually prove more fruitful.

The history of science offers plenty of reasons for caution about promises of rapid, predictable progress. The ambition to produce a particular technological result is not the same thing as understanding the scientific principles required to achieve it.

The alternative is to distribute funding across a wide range of research proposals, allowing scientific investigators considerable freedom to pursue their own questions. This reduces the danger of committing everything to a mistaken priority. It also preserves the diversity of research from which unexpected discoveries can emerge.

But distributing resources too thinly creates a different problem. Research often requires sustained investment in people, equipment, laboratories and experiments. A promising idea may never mature if it receives only enough funding to remain alive. And when funding is spread across too many projects, researchers may struggle to assemble the resources needed to make substantial progress.

Neither strategy eliminates the need to make choices, however. Even a funding system that aims to support the widest possible range of meritorious research must decide what counts as merit, which proposals deserve priority and how much support each should receive.

There is a further difficulty. Scientific reputations, institutional prestige and established research fashions can influence these decisions. Researchers working in recognised fields and well-connected institutions may be easier for funding agencies to identify. Meanwhile, an original idea may be difficult to evaluate precisely because it does not fit comfortably into an established category.

This is where a group of researchers has proposed a fascinating possibility: perhaps we can learn to recognise the early signs of scientific breakthroughs by studying the development of previous ones.^[1]

Can a breakthrough be detected before it becomes famous?

In a study of biomedical research, a team of researchers attempted to identify patterns that appeared before transformative scientific discoveries received widespread recognition.

Their starting point was a large collection of scientific publications. By grouping roughly 18 million biomedical papers into research topics, they could examine how different fields developed over time. They then studied the earlier histories of 21 discoveries that eventually received major scientific awards, asking what those research areas looked like before the discoveries became celebrated.^[1]

The researchers found a recurring pattern. Before a breakthrough, a field often experiences a burst of new research activity. Scientists from other areas begin moving into it. New publications appear, and some rapidly attract attention from other researchers.

The team developed a model using several features of this activity, including the proportion of new papers, the movement of research from other topics into a field, and the influence of individual publications.

Their results suggested that these patterns could provide a signal of an emerging breakthrough years before its eventual recognition. In a historical test involving signals from 1994 to 1997, they identified 18 research topics; the authors reported that 17 could be traced forward to a breakthrough.

The examples included the development of super-resolution fluorescence microscopy, directed evolution of proteins and enzymes, and the sequencing of the human genome.

The idea is intuitive once one thinks about it. A scientific discovery rarely transforms a field simply because one person announces a result. Other researchers investigate it, challenge it, develop it, connect it to other work and build upon it. A new scientific possibility begins to attract a community.

In this sense, scientists collectively register their interest through their research activity. They devote time, resources and professional reputations to exploring an idea. Their publications and collaborations leave traces that can be studied. The researchers’ work suggests that some of those traces may be detectable before the significance of a discovery becomes obvious to the wider scientific community.

There are substantial limitations, of course. The study’s training set of celebrated breakthroughs was small, the definition of a breakthrough is not straightforward, and a promising research signal can lead to a dead end. The authors themselves emphasise that their model should not replace scientific judgement or be treated as a reliable prediction of every major discovery.

But the study raises a question with implications relevant to our concerns: if patterns of scientific activity can reveal emerging opportunities, what might they reveal about the scientific possibilities taking shape across Africa?

The first problem is knowing what is happening

A country cannot intelligently develop scientific capabilities it cannot adequately see. To understand why, consider the difference between a list of scientists and a map of a scientific system.

A list might tell us that a university has researchers in chemistry, biology, engineering, mathematics and computer science. A map would tell us which problems those researchers are working on, what equipment they have access to, how their work relates to research elsewhere, and where their expertise might complement that of other people.

Scientific research is increasingly interdisciplinary. A problem in environmental monitoring might require chemistry, materials science, electronics and data analysis. Developing a diagnostic technology might require biologists, clinicians, engineers and specialists in manufacturing. A university may possess several of these capabilities without any one department possessing all of them. A related capability may exist at another university entirely. The challenge is to make these connections discoverable.

Africa’s universities and research institutions already contain scientific work across a broad range of fields. The University of Nigeria, Nsukka, for example, identifies materials chemistry research involving materials for treating industrial effluents, corrosion inhibition, catalytic materials, renewable energy and greenhouse-gas mitigation. Its materials-chemistry research group is coordinated by Professor P. O. Ukoha.^[6]

These point towards problems that intersect with industrial production, environmental management, energy systems and the use of materials in chemical processes. But a research-group webpage does not tell us everything we would need to know. Which specific materials have been developed? What results have been demonstrated? What facilities are available? Which other African researchers are working on related problems? Are there companies that could test or use the resulting technologies? What prevents the work from progressing further?

These questions require a continuously updated picture of scientific activity and the ability to connect that information to the people doing the research.

A similar exercise could be conducted across African universities and research institutions: mapping research groups, publications, grants, laboratory capabilities, equipment, collaborations and areas of expertise. Over time, the result could be a much clearer picture of where scientific knowledge is accumulating, where related work is taking place and where opportunities for collaboration might exist.

Such a map would not tell us automatically which research is important. But it would help us discover work that might otherwise remain outside the attention of potential collaborators, funders and institutions. And that is already a meaningful scientific capability in its own right.

Three windows into African scientific possibility

We can take a look at research already being conducted on the continent to make the point clearer. These examples are not intended as evidence that an algorithm would necessarily have predicted the discoveries in advance. They are meant to illustrate the kinds of scientific activity, institutional relationships and emerging questions that a more connected research system could help identify and develop.

The Atlantic Niño: African research and the global climate system

Hyacinth C. Nnamchi, an atmospheric scientist affiliated with the University of Nigeria, Nsukka, has contributed to research on the Atlantic Niño, a recurring pattern of sea-surface temperature variability in the equatorial Atlantic.^[2]

The phenomenon is important to climate science because changes in the tropical Atlantic can influence atmospheric circulation and rainfall patterns across parts of Africa and beyond. Understanding its mechanisms is relevant to the broader problem of predicting climate variability.

In their 2015 paper in Nature Communications, Nnamchi and his collaborators investigated the physical processes governing the Atlantic Niño. Their numerical experiments challenged the prevailing emphasis on dynamic interactions between the ocean and atmosphere, finding that thermodynamic feedbacks could account for a substantial proportion of the observed interannual variability. The authors argued that existing theories of Atlantic climate variability and its predictability needed to be reconsidered.

A subsequent paper, published in 2021, examined the seasonal behaviour of the Atlantic Niño. Using observations and atmospheric analyses, the researchers found that diabatic heating associated with the seasonal migration of the Intertropical Convergence Zone plays a central role in governing its seasonality. Their results questioned how much the ocean’s response to atmospheric variability contributes to the phenomenon.^[3]

So we see research connected to a Nigerian university contributing to a debate about the mechanisms of a global climate system. The research also illustrates the importance of connections. Climate science depends on observations, numerical models, computing resources and expertise distributed across institutions and countries. A scientific map could help identify African researchers working on related problems in atmospheric physics, oceanography, climate modelling, hydrology and remote sensing.

It could also help reveal where collaborations or shared infrastructure might support new research. A researcher investigating rainfall variability, for example, might benefit from knowing about a group studying atmospheric processes over the tropical Atlantic. A university seeking to develop climate-modelling capacity might discover relevant expertise elsewhere on the continent.

The opportunity is not to assume that every connection will produce a breakthrough. It is to make useful connections easier to find, evaluate and pursue.

Colorectal cancer: what Nigerian patients can teach the world about cancer

A different example comes from the African Research Group for Oncology (ARGO), a collaborative research network established in 2013 involving Obafemi Awolowo University and Memorial Sloan Kettering Cancer Center, with participating hospitals across Nigeria.^[4]

The network was developed to undertake rigorous cancer research in Nigeria and strengthen research capacity through training and mentorship. Its work has included prospective clinical databases, patient follow-up, biological samples and molecular investigations.

One of its important contributions has been to investigate whether colorectal cancer in Nigerian patients differs from the disease described in populations that have dominated much of the international cancer literature.

In a study of molecular and clinical characteristics, researchers found that the proportion of colorectal cancers exhibiting high microsatellite instability in a Nigerian cohort was significantly higher than in comparison datasets from North American and international genomic research, including datasets containing African American patients.^[5]

This is a particularly revealing example of how scientific research can change the questions the wider scientific community asks. Research in Africa is not only about applying medical knowledge developed elsewhere to African patients. It can contribute to knowledge about cancer biology itself.

It also shows what a scientific network can accomplish. A clinical observation at one hospital can become part of a larger investigation when patient recruitment, standardised data collection, laboratory analysis and research expertise are organised across institutions. ARGO’s published account describes a network whose work expanded across Nigerian institutions and geopolitical zones while combining research with training and capacity-building.

The network’s subsequent development has included more than a dozen prospective cancer databases and studies involving hospitals across Nigeria’s geopolitical zones. That is a scientific capability built through institutional relationships, not simply through the work of one laboratory.^[4]

A scientific radar could help identify further questions emerging from such a network: which patient populations remain underrepresented, which molecular findings need independent investigation, and which clinical or laboratory capabilities would make new studies possible?

A research network can generate opportunities beyond the particular study that brought it together.

Materials chemistry: the possibilities inside a university research group

We might as well take a look at the earlier-mentioned materials-chemistry research at the University of Nigeria, Nsukka.

The university’s description of the research group coordinated by Professor P. O. Ukoha identifies work on materials for treating industrial effluents, corrosion inhibition, catalytic materials, fossil fuels, greenhouse-gas mitigation and renewable energy.^[6]

The subjects alone reveal a potentially important intersection of scientific and industrial questions. Treating industrial effluents involves chemistry and materials science, but its practical development may also require knowledge of industrial processes, environmental standards, manufacturing and the economics of deployment.

A material that works under laboratory conditions may need to be tested against actual industrial waste streams. A catalyst may require further work on durability, production costs and performance at scale. And a promising research result may need an engineering partner before it can become a usable process. A map of research capabilities could help identify these different requirements. It could reveal researchers in related fields, locate relevant testing facilities, and help universities and companies discover where their expertise might complement one another.

But this example also exposes the limits of what we currently know. A university’s description of a research group’s interests is not enough to establish which technologies have been demonstrated, how mature they are or whether industrial partners are available.

Those are precisely the questions that a serious investigation of Africa’s scientific landscape should answer. A scientific map should not simply advertise research interests. It should help distinguish proposed work from demonstrated results, early-stage research from validated technologies, and potential connections from collaborations that actually exist.

Taken together, these examples show three different dimensions of African scientific possibility: research that contributes to global understanding, research networks that generate new knowledge about African populations, and university-based expertise that could connect scientific discovery to industrial development.

What they share is that their future potential depends partly on the ability to recognise and develop connections that may not be obvious from any one paper, institution or researcher’s biography.

The danger of building a scientific radar that sees only what is already visible

There is, however, a serious problem with simply importing the original research team’s approach.

A model based on publications and citations will inevitably reflect the scientific activity represented in its data. Research fields with large numbers of papers, well-established journals and extensive international collaboration will be easier to measure than fields whose work is less visible in those databases.

That matters especially in African science.

A small research group may be pursuing a genuinely original question but lack the funding, equipment or personnel to produce a large volume of publications. A researcher may be working on a problem of considerable local importance that attracts limited international attention. Early-stage research may be taking place through informal collaborations before it produces a recognisable cluster of papers.

An algorithm that looks only for rapid publication growth and citation influence could overlook such work. It might identify the fields already receiving substantial attention while missing some of the scientific possibilities that deserve attention precisely because they have not yet attracted it.

This creates a danger of circularity. Researchers and institutions that are already visible attract more funding and collaboration. Their activity produces more publications, making them even easier to identify as important. Less visible researchers remain outside the networks through which scientific opportunities are recognised.

The result could be a system that mistakes visibility for scientific potential. This is why an African scientific radar would need to go beyond bibliometric data. It could combine publication and citation analysis with information about research grants, laboratory capabilities, equipment, researcher registries, institutional surveys and direct participation by scientists.

It would also need to recognise that different disciplines leave different kinds of evidence. A mathematician’s work may not generate the same publication patterns as a clinical research network. An engineering group developing a prototype may produce technical reports, patents or working systems that are not adequately represented in journal databases. And some of the most important scientific work may consist of foundational research whose eventual applications cannot yet be anticipated.

A useful system would therefore need to treat its signals as invitations to investigate, not as final judgements about which researchers or fields deserve support.

The goal should not be to rank African scientists by their predicted contribution to science. It should be to make scientific possibilities easier to discover and assess, especially when they fall outside established patterns of prestige and attention.

What should Africa do with such a map?

The practical implications extend beyond the allocation of research grants.

A national or continental map of scientific capabilities could help universities identify potential collaborators, help research agencies understand emerging fields, and help companies discover relevant expertise in institutions they might not ordinarily approach.

It could also help governments distinguish between different kinds of research needs. Some fields may require sustained support for basic science. Others may need shared equipment, clinical research infrastructure, computing facilities or access to specialised laboratories. Still others may need partnerships with industry to test whether scientific findings can be translated into practical technologies. These are different problems, and they require different interventions.

The same information could help identify gaps in the scientific system. If several universities have researchers working on related problems but lack access to a particular instrument, shared facilities may be more useful than distributing small grants to each group. If a promising field has few trained researchers, fellowships and training programmes may be more appropriate than immediately funding large technology-development projects.

But the existence of a scientific map would not solve the problem of funding itself. Nor would it eliminate political choices about which national priorities deserve public resources.

It could, however, make those choices better informed. It could reveal capabilities that were previously overlooked, challenge assumptions about which institutions possess relevant expertise, and help funding agencies understand what kinds of support different research fields require.

The most sensible approach would be to test these possibilities through a limited experiment rather than promise a technological solution to Africa’s scientific problems.

A group of universities or research agencies could begin by mapping a particular field, perhaps environmental materials science, agricultural biotechnology or climate research. Researchers could identify existing capabilities, emerging research topics, gaps in infrastructure and opportunities for collaboration.

The next step would be to support a small number of projects or collaborations identified through the process, while retaining a comparison group or other means of evaluating whether the intervention made a difference.

Did researchers discover useful collaborators? Did previously disconnected groups begin working together? Did shared access to equipment enable experiments that had not been possible before? Did the process reveal promising research that conventional funding arrangements had overlooked?

These are measurable questions. They would also help determine whether a more comprehensive scientific mapping system would justify the resources required to maintain it.

There would be failures and false positives. Some research opportunities would not develop. Some collaborations would produce little. Some fields would turn out to be less promising than initial evidence suggested. That is not a reason to abandon the idea. It is a reason to build a system capable of learning from its own decisions.

And any such system would need to preserve a substantial place for investigator-led research. The discoveries of tomorrow may emerge from questions that do not yet register as priorities, from fields that have not yet attracted large research communities, or from ideas whose eventual importance cannot be predicted.

A scientific radar should help a country see more of its scientific landscape, not restrict the landscape to what its instruments can already measure.

From a collection of scientists to a connected scientific system

The continent’s scientific development is often discussed in terms of what it lacks: funding, laboratories, equipment, industrial partners, research infrastructure and international visibility. These constraints are real. But a picture of African science drawn only from its deficits is incomplete.

There are also researchers doing work that contributes to global scientific knowledge, institutions accumulating expertise, and networks generating findings that could not have emerged from isolated investigations. The question is how to make these capabilities more visible to the people and institutions that could develop them.

That requires more than a directory of African scientists or a website celebrating research achievements. It requires a living picture of scientific activity: who is working on what, what they have demonstrated, which capabilities they possess, what they need, and where their work connects to the work of others.

It requires institutions that can use this knowledge without reducing science to a collection of measurable outputs. It requires funding systems that can support both emerging opportunities and the less predictable foundations of future discovery. And it requires researchers to be able to find one another across institutional, disciplinary and national boundaries.

As I will continue to repeat, the ambition is not simply to make African science more visible to the rest of the world. It is to make African science more visible to itself.

A scientist at one African university should be able to discover a colleague at another institution whose work could transform an experiment. A company should have a way to find the scientific capabilities relevant to a problem it is trying to solve. A research agency should be able to identify promising fields without relying exclusively on prestige, established connections or international fashions.

And a curious reader should be able to see that African science is not a distant collection of isolated achievements. It is a landscape of people, ideas, institutions and possibilities that can be explored, connected and developed.

Before we ask whether Africa is funding its next great scientific breakthrough, we should ask whether Africa can recognise where its next great scientific possibilities are taking shape.

The researchers are probably already here and the ideas have probably already emerged. The first task is to find them, understand what they are doing, and connect them to the people and institutions that can help them grow.