Questions of subjectivity and objectivity have been very controversial in today’s postmodernist era. The divide is something we can never bridge, as we are programmed to view even objective phenomena objectively, as evidenced by instinct blindness for example or the fact that some animals have a much wider colour spectrum (think about how small the visual field is in the electromagnetic spectrum!). Besides that, we will never be able to enter someone else’s subjectivity, or as Thomas Nagel puts it, we will never know what it feels like to be a bat.
I think the best bet we have for achieving objectivity is replicating studies, double-blinding tests, and using large sample sizes and truthful statistical parameters. The external validity is, in a sense, synonymous to the objectivity of your studies. With animal models, sample sizes will pose an issue due to financial and spatial constraints. In the Denny Lab, one of my main tasks was finding a way to streamline the cell counting process in whole brain images, which demanded many hours staring at images of mouse brains. Eventually, the cells that had been stained using immunohistochemistry, would be counted, the data gathered and then hypotheses tested. In that context, the cell number would have been an indicator for memory formation. In the context of the Axel Lab, we are looking at numbers of cells in the olftcory epithelium, in order to study sensitivity tio various odors upon fear conditioning as part of a larger epigenetics project. The technology used is very sophisticated- a 3D imaging software called IMARIS. As the findings risk being very contentious, my post doc is keen on having as many counts as possible in order to minimise bias and subjectivity. The technician and 3 undergraduates including myself counted the same cells and compared our results. Three of us had very close counts, which was indeed promising, but the other undergraduate had drastically different counts, which definitely caused anxiety. First I thought that there must be a difference in what is perceived as a cell, and then if we standardise what counts as a cell or not, it would solve the problem. But that begs the question “who determines what a cell is?”, pof course in some slices where the staining is clear and the resolution sharp, it is easier to use a standardised metric, however there will always be slices that are more difficult to decipher, and so I guess those would have to be discarded (which then rducesour sample size). Nevertheless, it feels unsettling to have to conform cell counts to an arbitrary standard, as doesn't that just re-introduce a bias that we were trying to avoid? This conundrum was a topic of conversation among the five of us as we stared at the cell count excel sheet in front of us, and it bothered some more than others.
What does this mean for animal model research? We will assuredly never know what it feels like to be a mouse, nor can we, as a matter of tautology, ever escape our subjective perceptions. My instinct is to continue doing the best science possible, minimising error and bias wherever possible, and making sure that through replication and the emerging frontiers of technology, external validity is achieved. I may be skeptical of findings, as we should all be, but conducting research is a wrestle between your ego and nature, and we have to do our best to subdue the ego, that is, subdue our personal investments in the workings of the brain, and surrender to nature.
IMARIS cell counting