Are you a lumper or a splitter? Both?

There are lumpers, who like to group things into broad categories, and there are splitters, who divide things into smaller categories. While both ideologies can be useful, lumping is more often overdone. People put things that don’t necessarily belong in the same category together, which can lead to great misunderstandings, particularly when it comes to science.

Lumping seems to be a particular problem when it comes to genetic engineering.

When a scientific study shows a certain effect (positive or negative) of a certain genetically engineered trait in a certain plant species on a certain animal species when it is consumed at certain dietary levels, it does not ever mean that all genetically engineered traits in all plant species have that effect on all animals.

Each trait is different, will act differently in different crop species and different farming systems, and could have different effects on different animal species. Rarely can we make broad lumping statements about science, including genetic engineering.

Strangely, splitting is also a problem when it comes to genetic engineering, even though it isn’t that much of a problem with science in general.

For example, there are concerns that herbicide resistance genes will flow into crop relatives. This is a real concern, but splitters here forget that herbicide resistance genes can be developed with breeding and mutagenesis as well with by genetic engineering. If you are against the spread of herbicide resistance genes, then you should be against all herbicide resistance genes, no matter how they were made. Here, lumping might be more appropriate than splitting.

Another splitting example that is a big pet peeve of mine is that of plant intellectual property. Patents on genes are a huge concern with genetically engineered crops, of that there is no doubt. Something feels inherently wrong about people owning a gene or patenting an organism. However, there are similar issues with other forms of intellectual property when it comes to plants.

With Plant Variety Protection, a breeder can control whether or not other people are allowed to use the protected varieties in their own breeding programs. They can take legal recourse if someone uses the protected varieties without permission. There are some allowances for farmers to save seed, but they are limited. Plant patents, used for plants that aren’t reproduced sexually, do not allow for farmers to save seeds. Seed from hybrid plants, due to their biology, can not be saved (if the seed from hybrids is planted, the resulting plants won’t look anything like their parents).

Some people reject all plant intellectual property and choose only open pollinated and otherwise unprotected plant varieties. This I respect as a consistent position. Many more people only reject plant intellectual property when it comes to genetic engineering and ignore all other forms of plant intellectual property (not to mention all other intellectual property law).

In order to really understand any problems or benefits of plant intellectual property, we have to look at how all of its various forms affect farmers and breeders by lumping, not splitting.

Knowing when to apply splitting or lumping isn’t easy, but has to be learned for us to gain better understanding of all science and specifically of genetic engineering.


  1. a little simple re attacking the splitting of herbicide resistance gene activities. Are these breeding for resistance activities very widespread? Are there enough to bring them into the argument?

    It seems appropriate to tackle the primary impactor(s).

    Also, many might think that selective breeding is a safer method to allow some consideration for unknown complexities. I don’t know if that’s accurate.

  2. "Also, many might think that selective breeding is a safer method to allow some consideration for unknown complexities. I don’t know if that’s accurate"

    Far from avoiding unknown complexities, selective breeding leads to far more unknown complexities than genetic modification.

    When you breed in your herbicide/disease resistance trait, by the nature of genetics you lose a random selection of beneficial disease resistance, yield, grain quality, plant height, nutrient use efficiency, photosynthetic capacity, structural, root morphology etc etc etc traits. OK by laborious selective breeding over many years you can save as many of these as possible, but thousands of contributing factors accounting for a successful variety remain unknown and will be lost.

    Despite any perceived unintended side effects of GMOs (very few have actually been demonstrated) a far safer system is to genetically engineer your trait into a range of already successful and widely used crop varieties.

    Contrary to popular argument, I believe that GM has the potential to increase the range of crop varieties under cultivation if accepted as a technology. A myriad of high quality varieties that have been lost over the years when they became susceptible to a new strain of disease or had too low a level of water use/nutrient uptake now have the possibility to be resurrected by genetic engineering techniques and the incorporation of traits from other varieties that would protect their ‘achilles heel’.

  3. James, creating herbicide resistance via breeding and mutagenisis haven’t been as successful as creating glyphosate resistance with genetic engineering. There are, however, non-GE herbicide resistant crops in use. DuPont sells crops bred to have resistance to sulfonylureas. There are others, but I must be off to do homework 🙂

    Side note before I go – have you heard of resistance to glyphosate developing in South American coca plants? What’s happening there is effectively breeding for herbicide resistance. If it can be done accidentally, couldn’t it be done on purpose? And, it has, particularly in mutagenesis screening experiments.

Comments are closed.