Do we see through a microscope?

I would have considered this to be a question that barely needs an answer. That, however, was before I read an essay by philosopher Ian Hacking who brings the spirit of logical positivism to the question. When we look through a microscope we see objects that are not “… physical things in a literal sense,”, he writes, “but merely by courtesy of language and pictorial imagination.” Confused? Read on.

First of all, “seeing” is not the same as “looking”.   “Seeing” implies a higher level activity than mere “looking”.   Both involve the reaction of light on the retina, leading to stimulation of the optical nerve.   However, what we see when we look down a microscope differs from what we see when we look at a view out of a window in two important ways. The first is that most of the time our eyes are analysing reflected light, whereas the microscopist mostly (though not always) is looking at transmitted light, light that has passed through the object we are looking at.   The tones we see are a consequence of variations in the proportions of light that are transmitted or absorbed rather than in the proportions that are reflected or absorbed.

The second difference is that a consequence of magnification using an optical microscope is a very shallow depth of field (i.e. the space within which objects are in acceptable focus).   The picture below shows the view down my microscope as I was examining the specimens that I wrote about in my previous post.   In practice, I never actually “saw” this image, and I certainly could not have photographed it, due to the limitations of depth of field. In practice, I was constantly adjusting the focus with my right hand, whilst drawing with my left.   This is a point that Hacking makes: you have to learn to “see” through a microscope. As children we learn to judge distance by correlating visual information with tactile experiences, so microscopists learn to infer three dimensions in their worlds by gentle manipulation of the fine focus mechanism on their microscopes.   “Seeing”, in other words, is far more than just the consequence of light hitting the retina.


River Ehen, near the outflow from Ennerdale Water, December 2015. Drawn at x400 magnification.   See images in previous post for indications of scale.

More broadly, we could argue that “seeing” in the microscopic world requires interventions.   The lenses in our microscopes are one type of intervention, of course and without these we would not be able to see tiny objects with clarity.   But this is just the start: we often use stains to adjust the optical properties of certain parts of cells, making it both easier to see these and, in many cases, to infer their composition (iodine binding to starch is a good example).   These stains are often highly toxic so one unwelcome “intervention” is that we necessarily study dead or dying tissues.   The study of diatoms depends upon high resolution mountants, which make the task of viewing structural details of objects that are, essentially, microsocopic lumps of glass mounted between a glass slide and a glass cover slip much easier. This follows a stage in which the cellular contents are dissolved away by the action of strong oxidising agents.   So routine is this stage that many diatomists are in danger of forgetting that the empty glass shell is not really the reality that they seek.

We could argue that microscopic observations approach truth only when they can be validated by an independent means. At low magnifications, the difference between the raw and magnified images are small and individuals can easily validate what they see through a lens.   I know that there are words on a page of writing because I can see them crisply through my reading glasses but I could also, if necessary, hand the page to my son who will be able to confirm that the fuzzy blur that my unaided eyes discern is, indeed, writing.   However, there is a threshold in magnification beyond which this form of validation will not work.   We rely on a network of theoretical “proofs” to confirm the truth in a microscopic image. At one end of the scale this may involve a continuum of consistent observations from the hand lens to medium or high power optical microscope.   At the other end of the scale, it may involve a completely different approach to visualisation (the electron microscope, for example) to corroborate an observation. During my postgraduate days, a fellow student defined the cellular organelles he saw in his electron micrographs as “reproducible artefacts”.   His flippancy, in fact, contains a deep truth: we make our observations on highly manipulated specimens and need to be rigorous in our efforts to confirm their validity.

Hacking’s conclusion was that we see with a microscope, but it is naïve to suggest that we see through a microscope. We may look through a microscope but the images make sense only because of interactions that go beyond patterns of light hitting the retina.   “Seeing”, in other words, is a higher-level process that involves interactions with the brain and with the specimen itself.   Hacking calls on a lot of high power philosophy to make his point; all we need to do is to take care before equating the images we create with the truth.

Happy Christmas.


River Wear, Wolsingham: four views down the microscope: top left: January 2009; top right: March 2009; bottom left: June 2009; bottom right: September 2009.  


Ian Hacking (1981). Do we see through a microscope? Pacific Philosophical Quarterly 62: 305-322.