GM-free Scotland

A comment on science as we know it.

The Editor of the respected British Food Journal (BFJ) notes that because “research methodology concerns the unknown, hypotheses, probability, balancing and judging evidence or data”, science and research are not about facts.

This editorial comment arose after the Journal was made aware of a possible problem with the validity of a paper it had published in 2003. The study described in this paper claimed to have evidence that consumers preferred GM sweetcorn to conventional sweetcorn in one farm store based on finding that sales of the GM variety were 50% higher. Photographs taken in the store at the time indicated that signs above the two types of corn were not as described in the paper, but were worded so as to bias customer choice (see MISCONCIEVED WORMS and SLAPP ON FREE SPEECH – News, November 2007).

A scientific ethics researcher, who studies research conducted at the University of Cambridge, said the signs were “grounds for the journal to retract the article”.

Canadian Professor, Joe Cumins, also asked the British Food Journal to withdraw the article on the grounds that “the cornerstone of science is full and honest reporting, and this experiment and its controls do not appear to have been reported either fully or honestly.”

However, the Journal is standing by its decision to accept the paper, and by its award for excellence given the following year.

OUR COMMENT

Is it surprising that a leading journal is so unconcerned on learning that the data presented in one of its papers may not be what it seems?

The BFJ had clear clues of a possible bias in the findings. The extent of the PR exercise which preceded and accompanied all stages of data collection for the study is described in the paper: in the interests of “open communications”, the print and broadcast media were called upon for coverage, letters and public meetings were arranged locally, extensive pamphlets, posters and bookmarks were on offer, educational walks around the farm were made available, and all topped up by an interactive website with video footage specially developed for the occasion. The actual contents of these communications is not described, but there is a potential during each and every one of them for consumer bias.

The sources of funding for the study are, of course, detailed in the paper: two biotech industry support organisations, two commercial interest groups and a known pro-GM government department. All four authors of the paper are intimately involved with the Food Safety Network, which is heavily funded by the biotech industry. (At the time of the experiment the Food Safety was centered in the University of Guelph along with the authors, but is in fact the baby of the lead author and has since moved, with him, to Kansas University). Add to this the more general reality faced by today's researchers who always have to keep one eye on the funding for the next research project, and are inevitably under pressure to ensure their research is industry-friendly.

As commented on GMfreescotland.net, the extent of data-less conclusions voiced in the paper is breathtaking.

To go back to the BFJ Editor's comments. Science and research which have been bent by PR, 'corrected' to suit the funding bodies, and lacking honestly-described data are, indeed, not likely to be about facts. But surely “balancing and judging” the “evidence or data” includes consideration of the extent to which conclusions may have been biased by such factors? What is the role of leading scientific journals if not to minimise publication of papers whose integrity is compromised?

The requests to withdraw the paper were made because of signs in place during the data collection period describing the corn the authors didn't want customers to buy as 'wormy' and describing the corn the authors did want customers to buy as 'quality'. However, if the British Food Journal was prepared to overlook three major, and obvious, sources of bias, the presence of a couple of signs becomes a drop in the ocean of the bias already there.

A rather more sinister aspect to this story is that researchers who have revealed negative findings about GM have been treated very differently. Chapela and Quist, who published evidence of GM contamination in Mexican maize landraces, had their paper withdrawn by the journal: their conclusions were questioned because the specificity of their methodology in one part was considered inadequate (i.e. there was insufficient data for one of the conclusions voiced). Dr. Arpad Pusztai was black-listed and prevented from completing his research when he revealed on one occasion on radio that he had evidence of harmful effects from GM food (i.e. pre-publication PR). Dr. Irina Ermakova's evidence of multiple harm from a GM feeding study seems to have led a respected British journal to orchestrate a character assassination exercise on her (i.e. shoot-the-messenger).

Ask yourself this. Would the British Food Journal have been so keen to publish a small study funded by anti-GM bodies, surrounded by extensive anti-GM PR, and which concluded, on the basis of overstretched data, that non-GM food significantly outsold GM food, and that consumers were more concerned about harm from GM than from pesticides? Would the British Food Journal have refused to withdraw such a paper if there was subsequent evidence of signs calling the non-GM food 'wholesome' and the GM food 'unpredictably toxic', or, severe criticism of the methodology?

Perhaps the editors of scientific journals should remember that science can easily be confounded if the researchers don't really want to know the 'unknown', or, researchers frame their hypotheses so as to avoid finding out the 'unknown', or, researchers design their experiments so that they will only reveal what they want to be known.

Referring to the Editor of the British Food Journal, the Institute of Science in Society asked “Can we trust anything else published in the journal he edits?” But then, if science and research aren't about facts, can we trust anything else published in any scientific journal?

SOURCES

• British Food Journal 2003 106:10 pp.700-13 and 2006 108:8 Letters and Editorial
• www.freenetpages.co.uk/hp/a.pusztai/
• www.i-sis.org/hgt.php
• New Scientist 3.11.0
• Nature Biotechnology 2007 25:9 pp.981-7