Genetically Modified Crops: Science, Sociology and Statistics


  • Author: Chris Smaje
  • Date: 27 Nov 2014
  • Copyright: Image appears courtesy of iStock Photo

The use of genetically modified crops has become one of those touchstone issues that seems to divide people into two mutually uncomprehending camps. A disclosure: I’m on record as an opponent of GM crops, generally speaking, which will no doubt become clear from my remarks below. But my principal aim in this article isn’t to set out that particular stall. Instead, in a forum called Statistics Views, I want to explain why ‘views’ are at the heart of this debate – and indeed probably at the heart of any debate in which new technologies are hitched to social ends. That is why, as I shall argue, sociologists and social statisticians have a role to play in a field that might otherwise seem to fall within the bailiwick of the so-called ‘hard’ sciences.

thumbnail image: Genetically Modified Crops: Science, Sociology and Statistics

Indeed ‘science’ is a keyword in the GM debate, with GM critics often being accused of being anti-science by the technology’s proponents [1]. One of the contributions sociology has made is in finding meta levels of analysis that can reveal the hidden work of persuasion done by the concepts and metaphors people use in their disputes. Here I want to turn some sociological attention of this sort onto the concept of ‘science’ in the GM debate, and show how social statistics can help. The demonstrable success of science as a method of enquiry, and the foundation it establishes for new technological innovations, provides an incentive to stretch the referents of the term ‘science’ into arenas of policy and decision-making which are not fundamentally ‘scientific’. Sociological concepts of ideology can assist us here, and so can statistical analyses which reveal it quantitatively.

One of the first people to provide a systematic account of ideology was Francis Bacon (1561-1626) in his pioneering efforts to formulate the modern scientific method. Bacon identified ‘idols of the mind’ that interfered with correct scientific inference, and in so doing established the modern commonplace that distinguishes ‘truth’ from mere ideology [2]. But sociologists typically see all truth claims, including scientific ones, as inherently ideological in the sense that they frame the object of enquiry in particular ways, which are never the only ones possible. The power of the scientific method does not rest principally in the fact that scientific claims are non-ideological in this latter sense, but in the fact that they are subjected to rigorous and continual trial by falsification. This power, however, can easily be objectified into the notion that science establishes meta-ideological truth claims, and this in turn makes it ripe for appropriation for more clearly partisan ends.

Robert Paarlberg’s book Starved for Science [3], which suggests that European opposition to GM crops has led to starvation amongst Africa’s poor, is one example of how the word ‘science’ can lose its proper moorings in the scientific method and be used metaphorically to connote a particular type of agriculture or development package, involving a range of social, political and economic commitments that figure implicitly beneath the explicit commitment to science. The word ‘technology’ can be used in much the same way, as for example in Berezow and Campbell’s statement that “Instead of embracing technological progress, such as genetically modified crops, progressives have spread fear and misinformation” [4]. Like the ‘science’ of Paarlberg’s title, ‘technological progress’ is proffered here as an uncontestable benefit rather than as a particular and contestable social ideology of historical development. As Scoones has argued, “technologies are always linked to social, economic and political contexts, and the assumption that nothing else matters beyond the technical fix is of course deeply flawed” [5]. In this sense, I would argue that much of the opposition to GM crops is ‘pro-social context’ rather than ‘anti-science’ as such, though it does tend to oppose the use of the term ‘science’ as a tactic of closure to political debate, as in the examples above. The sociological concept of ideology, freed from its Baconian shackles, can help us distinguish science from such ‘scientism’, and to analyse what Kinchy has aptly called ‘scientized politics’ whereby advocates and opponents of GM technologies engage scientific literatures in complex ways, in her words “using and producing scientific knowledge as well as treating scientific discussions as a stage for launching complex social critiques” [6].

So much for sociology – what about statistics? Let me try to sketch a role for statistical analysis in service of this sociological critique of ideology, beginning with reference to a recent GM controversy, the so-called Séralini affair.

Gilles-Éric Séralini is a molecular biologist at the University of Caen whose team published a study in the academic journal Food and Chemical Toxicology that identified elevated morbidity and mortality among rats fed both on a GM maize variety NK603 tolerant to the herbicide Roundup and to low doses of Roundup itself. The paper was criticised on various grounds, including the inferences it made from its statistical analysis. Food and Chemical Toxicology subsequently undertook a review process, itself criticised for its opacity and inconsistency, as a result of which it retracted the paper. Later, the paper was republished in another journal, Environmental Sciences Europe [7].

I will leave judgments about the paper’s methodological shortcomings or otherwise to my biostatistical colleagues, focusing instead on some of the wider sociological lessons. Both proponents and opponents of GM technology are wont to impute various orders of disrepute to their antagonists, such as cherry-picking data and forming shadowy conspiracies of influence [8]. Again I don’t propose to assess such claims here, but it strikes me that the furore over papers in peer reviewed scientific journals arises in part because of the cachet of academia in general and science in particular of being above such low motivations. In sociological terms, a peer reviewed paper is a status good whose high standing accrues because it cannot be bought, traded or otherwise traduced. But it can then be used as a currency in other status battles – hence the bitter contestations over the gatekeeping of the peer review process in cases such as Séralini’s, and the debates over the balance of the evidence. Most of us who have experienced the reality of the peer review process would, I suspect, accept that it’s a more capricious thing than this lofty public image admits, which cannot really bear the weight of the public need for uncorrupted truth. This is the more so because academic funding and academic research agendas are increasingly bound up with public policy and private enterprise, as Dick Russell argued some years ago,

A university scientist with direct or indirect commercial interests simultaneously may be serving on government granting panels, testifying at Congressional hearings, publicly discussing the risks and benefits of new products….Institutions become overdependent on money from a single large corporation, and professors distracted from their proper duties [9]

It is here that social statistics applied as meta-analysis to the research process itself can be of help. A paper by Meyer and Hillman examined the three rat-feeding studies of NK603 maize and argued that all of them satisfy or fail to satisfy the evaluation criteria laid down by the European Food Safety Authority to a comparable extent, arguing that “the rejection of only one of the papers is, thus, not scientifically justified” [10]. More generally, a study by Diels et al reports statistically significant correlations between author affiliation to the GM industry and study results favourable to GM crops [11]. I am not so naïve as to suppose that these studies themselves will be regarded as being beyond methodological reproach, that the authors’ motives will not themselves be impugned, or that such analyses alone can restore the elusive cachet of objectivity to the research process. However, the accumulation of careful statistical meta-analysis of research results can surely help in time to clarify the balance of evidence more persuasively than the accumulation of opinions by scientists or indeed non-scientists about what ‘the science says’. And perhaps it can also help focus debate around the different evidentiary criteria typically deployed in the various relevant disciplines (agronomic, ecological, toxicological, microbiological and biomedical) which appear to have formed one part of the disputes around the Séralini study.

The degree of passion aroused over the Séralini affair seems curious if the original paper is considered in itself on narrowly scientific grounds, for whether the results prove robust in the long term or not they don’t necessarily carry wider implications for biotech deployments beyond NK603 maize. The same passions are ignited in other GM arenas, such as Vitamin A fortified golden rice, or in Paarlberg’s claims about GM crops and Africa’s poor. Aside from the disciplinary boundary disputes and the gatekeeping over peer review that I mentioned above, I’d suggest that the sociological root of these passions lies in strongly divergent political convictions about how best to secure human livelihood which are ‘ideological’ in the sociological sense I described earlier. Crudely put, perhaps these ideologies could be described on the one hand as an ideology of ‘technological progress’ of the kind championed by Berezow and Campbell, which invokes the power of science as a technique sufficient in itself to improve human wellbeing (or even to improve humanity, as in the project of transhumanism), but is not in itself ‘scientific’. And on the other as ideologies of specific and human-centred orientations to self-realisation in which science and technology at best can only ever serve these primary goals, and in some cases actively hinder them. When it comes to the practical deployment of GM technologies, doubtless one can adopt various intermediate positions that are neither wholly in favour nor against (this crop, for these people, in this situation). My argument here is simply that there seem to be deep ideological roots to the debate.

Renegade environmentalist Mark Lynas has adopted the discourse of climate change denial to describe anti-GM positions as ‘denialist’ [12]. It seems an inappropriate borrowing. Nobody is denying the scientific basis of GM technologies: what is at stake are the social consequences, and here we primarily have views to contend with – ideologies in the sociological rather than Baconian sense – which are, precisely, ‘views’ and therefore cannot be denied in the name of science, even if they can sometimes be illuminated by evidence from the social sciences [13]. But sociologists can contribute to the GM debate by elucidating the ideological basis of those views, and statisticians can contribute by subjecting claims concerning the scientific and evidential basis emerging from them to quantitative scrutiny.


1. Lynas, M. (2013) Lecture to Oxford Farming Conference.

2. Larrain, J. (1979) The Concept of Ideology, Hutchinson, New York.

3. Paarlberg, R. (2009) Starved for Science, Harvard, Cambridge, MA.

4. Berezow, A. and Campbell, H. (2013) Lefty nonsense: when progressives wage war on reason. New Scientist

5. Scoones, I. (2011) GM crops 10 years on.

6. Kinchy, A. (2010) Anti-genetic engineering activism and scientized politics in the case of ‘‘contaminated’’ Mexican maize. Agriculture and Human Values, 27: 505-17.

7. Séralini, G. et al (2014) Republished study: long-term toxicity of a Roundup herbicide and a Roundup-tolerant genetically modified maize. Environmental Sciences Europe. doi:10.1186/s12302-014-0014-5

8. See, for example, Philpott, T. (2012) The making of an agribusiness apologist. Mother Jones ; Entine, J. (2014) The debate about GMO safety is over. Forbes Magazine,

9. Quoted in Kloppenburg, J. (2004) First the Seed, University of Wisconsin Press, Madison, pp.198-199.

10. Meyer, H. and Hillbeck, A. (2014) Rat feeding studies with genetically modified maize - a comparative evaluation of applied methods and risk assessment standards. Environmental Sciences Europe. doi:10.1186/2190-4715-25-33.

11. Diels, J. et al (2011) Association of financial or professional conflict of interest to research outcomes on health risks or nutritional assessment studies of genetically modified products. Food Policy. doi: 10.1016/j.foodpol.2010.11.016


13. I’ve examined some of the relevant issues in relation to this sort of evidence in ‘Social statistics, counterfactuals and the green revolution’

Related Topics

Related Publications

Related Content

Site Footer


This website is provided by John Wiley & Sons Limited, The Atrium, Southern Gate, Chichester, West Sussex PO19 8SQ (Company No: 00641132, VAT No: 376766987)

Published features on are checked for statistical accuracy by a panel from the European Network for Business and Industrial Statistics (ENBIS)   to whom Wiley and express their gratitude. This panel are: Ron Kenett, David Steinberg, Shirley Coleman, Irena Ograjenšek, Fabrizio Ruggeri, Rainer Göb, Philippe Castagliola, Xavier Tort-Martorell, Bart De Ketelaere, Antonio Pievatolo, Martina Vandebroek, Lance Mitchell, Gilbert Saporta, Helmut Waldl and Stelios Psarakis.