No matter their agenda, this selection of books inevitably distorted what became accepted Knowledge for agronomists in Peru. However, the second layer of distortion is perhaps the most pernicious, and focuses on the detail of legitimate practice.

Many of these books championed the industrialisation of agriculture. This was perceived as ‘progress’ and assumed to lead to the greater ‘efficiency’ of agriculture. For example, the use of tractors over peasant labour was heavily promoted. This, the books argued, created greater ‘productivity’. This idea had real world consequences, when peasant farmers were displaced by machinery. In a culture where peasant families still knew which land they were supposed to be farming from Incan times, the personal impact was devastating. In addition, such extensive farming damages the soil through heavy use of fertilizers and pesticides, and the mechanical ploughing of topsoil breaks down its structure.

But let us ask what does greater productivity mean? The uninformed may assume that greater productivity means more produce and so cheaper costs of produce. But this is not what it means at all. The definition of productivity being imposed was how much produce could be produced per person working in agriculture. Not how much produce can be produced per hectare. The consequence of the ‘Knowledge’ was that peasants lost their livelihoods and had little choice but to migrate to shanty towns while up to 14 times less produce was actually produced per hectare (George, 1976). This, of course, matters little to the owner of the land, since they increase their profits at our expense and the inevitable human costs to other people. In addition, agriculturists were explicitly advised to target their agricultural produce at the middle class who could afford higher prices, or for export to the wealthy nations. Here, ‘science’ in the service of profit is very far from a neutral endeavour, and promoted nonsense ideas like ‘higher productivity’ must be better, and extensive farming means we are ‘all in the same family’ (this is a direct quote from one of the agronomists). Certainly, if we redefined productivity to mean how much is produced per hectare, science would be useful for all. In this real world situation, it is only useful for the wealthy. To understand how unscientific a methodology is, we need to expose the assumptions that lie beneath the science. In this case, for example, that agricultural land is owned by individuals rather than collectively. That those who own the land want to maximise profit rather than produce. That agriculture is a scientific industrial process, rather than a way of living for a community.

So what does this have to do with the ‘science of reading’? Similar distortions creep into many layers of the process.

We can identify these as:

1: University structures and funding
2: What counts as ‘research’ and ‘evidence’.
3: Ignoring the contradictions to support an ideological position

1: University structures and funding

The funding of universities, certainly in the ‘west', has become primarily dependent on published research. Lecturers must publish regularly or face an uncertain career. This has two direct consequences. The first is that it becomes increasingly advantageous to generate small scale research that costs less to undertake and leads more quickly to publication. The second is that the competition to get published means that papers that buck the current trends and fashion usually get refused. These both have a chilling effect on research. Indeed, it has been argued that Einstein would not have been able to publish his seminal scientific papers due to a lack of references (e.g. Daston, 2022).

In terms of research about reading, it is surely self-evident that small scale research on easily measurable aspects are much easier, than large scale research with complex evidence to gather. For example, it is very much easier to design a research project that focuses on an intervention to improve decoding, than a research project measuring interventions to improve comprehension. The first is easy to design, implement and measure. If we also argue that decoding is necessary to become an effective reader, then we can make grandiose claims about the research outcomes, In contrast, there are so many factors that impact on or lead to comprehension (such as prior knowledge, energy levels, use of metacognitive reading strategies, interest of lack of it in the content being read, anxiety, health and nutrition, motivation, availability of reading books in their actual lives, practice, etc), that measuring the impact of an intervention is hugely problematic. It may matter more than decoding, but you don’t make it easy on yourself to go down that route. A complex research project also makes it very difficult to isolate factors and arrive at robust conclusions. It also opens up the researcher to more challenges about their methodology. When your livelihood, or your department’s existence depends on published research, it is little wonder that researchers stick when possible to the most manageable options.

2: What counts as ‘research’ and ‘evidence’.

You will frequently read advocates of the ‘science of reading’ arguing that nothing counts as evidence unless it is supported by double-blind tests. This is hugely important in medical research where interventions may have a minimal positive impact while placebo may have an equivalent positive impact. Separating out placebo from ‘real’ medical intervention is evidently extremely important, particularly as placebo effects tend to diminish over time. However, it has less relevance in education.

First of all, it is almost impossible to set up research where the learner is unaware whether they are getting the ‘real’ intervention, or the placebo. The alternative intervention must have credibility to be convincing. In many cases, this means that the ‘placebo’ intervention can be expected to have a positive impact (because if it doesn’t, learners will see through it) and you will have troubling ethical considerations if you provide an intervention of no value), but that means you are not measuring the impact of the intervention, but the difference between the two. Secondly, how do you match the placebo group with the intervention group? Traditionally, psychologists have tended to use age, gender, and ‘IQ scores’. Leaving aside the complete nonsense that, if there is such a thing as a meaningful single measure of ‘intelligence’, the current ‘IQ tests’ can measure it, and ignoring the fact that visual tasks can be done verbally, and verbal tasks visually (Howe, 1997), what about all the other factors involved? How about social class? Or neurodivergence? Or how nurtured the individuals have been? After all we are not simply interested in biology in education.

Nevertheless, this argument that only evidence gathered from double blind research counts is hugely influential and used to dismiss all individual experiences of learning and the teachers’ and learners’ lived experiences. But there is another important difference between medicine and education. Interventions can, and in my view should, have much larger effect sizes, making double-blind tests sometimes redundant. Even in medical research, if a medical intervention has a much bigger impact on health than the placebo, researchers will often terminate the research and offer the intervention to those who had only had the placebo (for example: Gray et al, 2007). Should double-blind research, with all its methodological complications and compromises trump experience? Whose interest is this serving?

Creating large-scale double-blind research is primarily the preserve of wealthy universities. Should universities have the final word on what counts as research and legitimate evidence? What universities most value in their academic staff is publication, despite the fact that very few people including academics actually read most of it (Lattier, 2016). One of the ways academics demonstrate value, is by critiquing other academics. In my experience, they are very good at being critical, and relatively poor at providing solutions to problems. There is very little value seen in that.

3: Ignoring the contradictions and flaws to support an ideological position

I reported earlier, that the scientific approach to agronomy education failed to take into account important evidence. This is partly because, like most research, the focus is limited to make the research possible. Consequently, it may well be that agronomists writing those textbooks were simply unaware of the evidence they were ignoring. However, it is very easy to identify if you look for it.

Mary-Anne Wolfe, one of the more empathetic researchers in the field of dyslexia (perhaps because dyslexia runs in her family) wrote a very interesting and influential book called ‘Proust and the Squid’. In it, she described brain activity, measured through Magnetic Imaging, of the process of reading in milliseconds. One of the most interesting aspects of that description was the recognition that the reader understood what they were reading before they subvocalised the words. To my mind, this scientific evidence is crucially important. If people understand before sub-vocalising, what is the purpose of sub-vocalising? It has been suggested that it plays the role of checking for errors in the reading process. I might suggest that it is a product of being taught to ‘say’ the words as the way to achieve comprehension (not that that is always effective). In other words, it is a product of teaching, rather than necessary for reading. It turns out that it is quite easy to increase reading for meaning speeds dramatically while loosening the addiction to sub-vocalisation. So where does that leave phonics? Why is this important piece of evidence being ignored? I would suggest because the ‘science of reading’ is an idealogical position, not a scientific one. So unless an advocate of the science of reading can explain this evidence, why would we accept anything they claim? Credibility is, afterall, a two way street.

References:

Daston (2022) in Why Einstein Wouldn’t Be Published Today: A Conversation with Lorraine Daston, Part Two (Accessed 10.11.23)

George, S. (1976), How the Other Half Die, Penguin.

Gray R, Kigozi G, Serwadda D, Makumbi F, Watya S, Nalugoda F, Kiwanuka N, Moulton LH, Chaudhary MA, Chen MZ, Sewankambo NK, Wabwire-Mangen F, Bacon MC, Williams CF, Opendi P, Reynolds SJ, Laeyendecker O, Quinn TC, Wawer MJ, (2007) Male circumcision for HIV prevention in men in Rakai, Uganda: a randomised trial. Lancet. 2007 Feb 24;369(9562):657-66. doi: 10.1016/S0140-6736(07)60313-4

Howe, M. (1997), IQ in Question: The Truth about Intelligence, Sage.

Lattier D, Academics Write Rubbish Nobody Reads, FEE Stories (accessed 10.11.23)

Wolfe, M., (2007), Proust and the Squid, Harper Collins.