Notes from #OR2016 opening keynote - Laura Czerniewicz

7 min read

(Slides are available - note that these are hugely detailed and referenced, and you should take time to read through them. I should note that Laura was clear that her usage of terms like the "global south" as shorthand - which I have repeated in these notes - hid a great deal of sub-national inequality and were themseleves problematic)

Knowledge productuion and dissemeination has always been fraught with inequality. The global south is poorly represented, for instance, in Web of Science. Most citations are of academics residing in US and EU cites.

The global south is imaginary - Laura introduced the Brandt line as a way of visualising it, but inequality does not produce a simple dichotomy. Inequalities happen within spaces (this was illustrated by a view of two areas of Cape Town).

But what causes inequalities in knowledge production? Funding (as a % of GDP devoted to R&D) is very variable - but again is hugely concentrated in the global north, and China. But it is not just the money, we need also to look at legitamacy, reward systems and gatekeeping. The measurment of national representation in WoS itself has a bias against reports, consultancy and presentation.

Typologies and genres of research have different values in different systems - for example consultancy research is disparaged in Africa but is often the only available source of funding. In SA funding is linked to publication in (TR ISI) journals. Many institutions give money directly to authors - Cape Town does not.

Citations are a measure of credibility, but have uneven geographies - there is a huge bias toward the global north. Altmetrics are being taken up very slowly.

There are huge issues with access to research in the developing world - with many academics reporting huge difficulties in getting to the research they need.

The #1 producer of publications (linked to GDP) in Africa is Zimbabwe - combines a low GDP with a great scolarly tradition.

Only 1% of total journal articles are published by academics are based in South Saharan Africa.

Empirical sciences academics are having to research the global north in order to get published, whereas in the north academics are able to publish about the south. In a study of two "african studies" journals, the percentage of african-based authors has decreased.

Journal boards are primarily from highly developed countries. Journals are aware of this imbalance - international is a shorthand for the global north.

All this means that local knowledge may not be available to others in similar conditions, and is a necessary contribution to global knowledge. The choice for African scholars is often between recognition and relevence.

The opportunities of a networked world drawing on the changes that the internet has made to networks has led to the Budapest Declaration on OA.

The research cycle has changed at every point - from conceptualisation to publication to citation to measures of impact. And it continues to changes, offering genuine opportunities to collapse distance, enable global collaboration. But digital merely affords open, it does not equal open. Digital rights and licensing, digitally mediated closed networks, all offer new layers of complexity. 

There is a danger that the information revolution could exacerbate socio-spaital disintegration. It cannot be assumed, for example, that elecricity, computers and bandwidth are available everywhere at all times.

The rise of mobile connectivity - with a massive increase in the use of smart phones - is an important counter-trend. But though devices are getting cheaper, the cost of data in developing countries is the real barrier.

Discoverability is a new currency - if it can't be found it doesn't exist. Visibility (thus connectivity, infrastructure) is a requirement for participation. Internet based rankings and metrics affect the way we participate in physical space.

Search engines are the primary way that content is found at all levels - which changes the notion of what a search engine does. As a surrogate expert, a co-producer of knowledge - the "invisiblity of search engines" hides the biases and assumptions built into obfuscated algorithms. The "collective intelligence of the web" sounds like a closed cycle - the popular becomes more popular.

Personalisation, at an invdividual level via previous searches and profiling, keeps the seeker after knowledge trapped in their own cultural assumptions. The rise of filter-bubbles makes wider knowledge less discoverable. And the search engine market is dominated by google, as is the browser market.

Open Access policies are growing - but these may also have global unintened consequences. OA is slanted towards the north, and may make it even more difficult to discover knowledge from research conducted in the global south.

Looking at global poverty - a Cape Town University investigation started from the idea that poverty is taken seriously and research on it would be clearly available. Looking at academic and non-academic experiences around nearly every part of the world making two google searches and two google scholar searches for "poverty alleviation", no results came from SA in google, one in scholar. The number one ranked result was wikipedia, of interest because some access to wikipedia is free and papers cited in wikipedia tend to have more policy elsewhere.

Adding South Africa to the terms, an article in Forestry Policy and Economics in a very expensive journal appears - but the link that google finds is in an institutional responses. Google scholar finds more articles in repositories, and in green OA journals, than from other sources.

Turning to climate change (which is, of course unequal in cause), the team conducted similar research in what was percieved to be a newer and more open field. The US is dominant, with China at number two, and the global north better represented.

On google scholar, no results were found from South Africa, Africa or any other developing countries for a search on "climate change" - authors were largely from US and UK. The top ranked paper appeared in five web locations, three being repositories. Largely technical reports were returned, and only two different journals - this may be because technical reports are a more accepted form of scholarship in the field.

adding South Africa to the search terms, the top ranked article was in Nature, but using inverted commas returned the South African Journal of Science. 

Editorial oversight in the field is again concentrated in the UK, US and China - even for research specifically considering South Africa. This finding suggest that the "gatekeepers" have a particular cultural background with implications for the field.

Academics tend to store their identities with dedicated services rather than own their own infrastructure - reporting that online presence takes time, money and expertise. Choices were shaped by resource constraints and perceptions of value of certain activity.

In closing Laura noted that online practice adds major complexities to global inequalities of power and resources, but OA is only meaningful only if everyone has the ability to participate. She compared the earlier science examples with the much more evenly distributed open source community.

(Questions focused on the need for collaboration, and a need to investigate new technological and business models that work for everyone. There was an interesting point made about the complicity of search engine providers in algorithmic bias - search engines are being taken on but this may need more co-ordinated effort.)