An Organisation’s Umwelt

The views and opinions expressed on this account are my own and do not reflect the official policy or position of my employer.  Any content provided is for informational purposes only and should not be considered or relied upon as professional advice.

“Every animal is enclosed within its own sensory bubble, perceiving but a tiny sliver of an immense world… It doesn’t care about other stimuli, and probably doesn’t know that they exist... Our Umwelt is all we know, and so we easily mistake it for all there is to know.  This is an illusion that every creature shares.”[i]  Ed Yong

In previous posts, we explored how risk teams can be limited by their "license to operate", or by topics outside the organisation's "Overton Window". In this post we'll discuss another challenge: when risks are invisible to the organisation - or event to the risk team itself.

Umwelt

The term Umwelt was coined by the zoologist Jakob von Uexkull. It refers to the part of an animal’s surroundings that it can sense and experience: its perceptual world.  Animals evolve to perceive only what they need to survive, not the full reality of their environment.

The same concept applies to organisations. Companies operate within a “perceptual bubble” shaped by their data, language, and assumptions.  If something falls outside this bubble, the organisation might be unable to notice or discuss it.  Risk teams play a key role in identifying when these blind spots could lead to organisational harm.

Human Vision as an Analogy

Understanding how human vision simplifies reality provides a useful analogy for how organisations process and interpret data. 

Visual Model vs Reality

We don’t process the world in high-resolution detail.  Instead, our brains create a model based on experience and update it only when we notice something unexpected.  Only a tiny area of our visual field has sharp, colourful detail.  The rest is blurry and mostly colourless, yet we feel like we see everything clearly.  The brain fills in the gaps.

We use data to validate our assumptions about the world.

Vision interpreted through language

How we name colours influences how we see them.  William Gladstone noticed that a lot of early writing did not use the word ‘blue’ to describe things that we often think of as blue today.  For instance, in Homer’s the Odyssey - set on, in, or around water - Homer describes the ocean as “wine-dark” and other inventive ways, but he never uses the word ‘blue’.  This is probably not just because Homer did not have the right word at his disposal; the lack of a specific word may have meant he was unable to see blue in the way that we do. In every language studied, blue appears to be the last colour to get its own name.[ii] 

The Himba language does not distinguish between green and blue, instead dividing the world into many shades of green.  In an experiment[iii], Himba participants were shown a circle of twelve coloured squares.  In the first test, eleven squares were coloured the same shade of green, and the twelfth was coloured blue.  In the second test, the twelfth square was coloured a different shade of green.  Himba participants struggled to spot the blue square but easily picked out the green.  Europeans showed the opposite pattern; they quickly identified the blue square but found it hard to spot the ‘odd’ green square.  A related study[iv] found that native Russian speakers, who have separate words for light and dark blue, could tell shades of blue apart much faster than English speakers.

Language doesn’t just describe reality, it shapes how we experience it.

The Brain’s Causality Bias

Sensory signals reach your brain at different speeds, but your brain automatically synchronises them.  This is because your brain is wired to assume cause and effect.  TV broadcasters exploit this phenomenon: as long as audio and video feeds are close enough together, your brain aligns them automatically.*

Our hard-wired assumptions affect how we interpret data.

Data Isn’t “Truth”

Organisations often treat metrics as if they perfectly represent reality.  But, like vision, data is:

• A simplification of reality,

• Influenced by in-built assumptions about the world,

• Shaped by language and definitions, and

• Prone to biases in interpretation (especially causal ones).

Even when metrics are initially recognised as flawed, that nuance often fades after a few reporting cycles.  Measures that were once accepted only as indicative soon solidify into unquestioned representatives of reality.  There is something about a number that encourages collective amnesia about its limitations, even when they were explicitly acknowledged at the start.  As historian of science Lorrain Daston put it, numerical figures “hover between the realism of the invented and discovered”[v]. 

Movie Review Example

Think of film ratings:

• Critics provide qualitative reviews plus a 1–5 score.

• IMDB gives an average score out of 10.

• Rotten Tomatoes further simplifies scoring to like/dislike.

Each scoring system simplifies the scoring approach in different ways and to different degrees. Numeric scoring systems are easier to aggregate, but lose nuance. Using multiple sources can offset their limitations, but, when picking a film, we often rely on just one.

Categorisation: A Hidden but Powerful Force

Modern organisations handle vast amounts of data, which must be categorised to be useful. However, categorisation is not a neutral process - it shapes how the organisation understands itself. Sometimes management is involved in the process of weighing the pros and cons of each categorisation choice and settling on the least bad aproach, but in my experience this is rare. Instead categorisation choices are often shaped at the outset by immediate regulatory requirements, or default system settings. What is rarer still is for organisations to attempt to document the assumptions and limitations underpinning categorisation choices.

Here's why it matters

• Simplification: Categorisation simplifies complex information, but it involves trade-offs between precision and practicality.

• Inflexibility: Once adopted, categorisation systems are hard to change. Over time, people start treat them as objective reality, making it harder to imagine an alternative.

• Conflicts: Different teams may prefer different categorisation methods, leading to inconsistencies across the company or compromises that don't work for everyone. Groups with high social capital can impose their preferred system on everyone else.

Why Risk Teams Should Care

An organisation’s Umwelt includes far more than data—it includes frameworks, definitions of success, documentation practices, and the mental models people use to understand the business.  It captures all the ways in which the organisation understands itself.

Risk teams should:

• Understand how categorisation influences decision-making.

• Identify where perceptual limitations distort decisions.

• Maintain a long memory of the compromises made in developing an organisation's Umwelt, including the categorisation decisions that sit behind it.

• Consider how alternative ways of seeing the world might produce different outcomes.

• Challenge outdated categories or flawed metrics that have become unquestioned reality.

• Identify scenarios where changes to the internal strategy or external risk environment could exacerbate limitations in the organisation’s umwelt.

---

Hopefully that has sparked some ideas.  I am hoping to start a discussion with this blog.  If you find it interesting, please share! If you would like to contribute or share feedback, please comment below or message me on LinkedIn!

---

[i] HOW ANIMALS PERCEIVE THE WORLD, The Atlantic, June 13 2022, Ed Yong, https://www.theatlantic.com/magazine/archive/2022/07/light-noise-pollution-animal-sensory-impact/638446/  I also recommend Ed Yong’s book “An Immense World”

[ii] The Secret Lives of Colour, 2016, Kassia St Clair, 978-1-473-83081-9

[iii] Mylonas, Dimitris & Caparos, Serge & Davidoff, Jules. (2022). Augmenting a colour lexicon. Humanities and Social Sciences Communications. 9. 10.1057/s41599-022-01045-3.

[iv] Russian blues reveal effects of language on color discrimination, PNAS Vol. 104 No.19 7780-7785, Jonathan Winawer, Nathan Witthoft, Michael C. Frank and Lera Boroditsky, May 8, 2007, https://doi.org/10.1073/pnas.0701644104

[v] Why statistics tend not only to describe the world but to change it, London Review of Books, Vol. 22 No. 8 · 13 April 2000, Lorrain Daston

* Below 45 milliseconds (sound early) or 125 milliseconds (sound late).