The need for a frame of reference
Unlike physics, organisational theory and business theory is currently at the stage that there is no agreed frame of reference we can use to apply organisational theories with confidence. In this article I will explain how mature scientific domains move beyond using methods for experimental research and simple pluralistic models, and gradually develop, or through a major breakthrough identify, the optimal frame of reference within that domain. The most notable example is physics, where Isaac Newton identified the proper frame of reference of mechanics which involves applying the Law of Conservation of Energy to any mechanical systems that are under analysis. Before then we wouldn't know what questions to ask, or what to measure in a mechanical system to capture and predict its behaviour. Even if we already knew how to design experiments, we wouldn't know which experiments to try, and in what order. In this scenario, overall progress is often sporadic, piecemeal and uncertain.
Credit: James Robert White, Title: Study, Leaf
The current state of data science for business
In data science for business we already have methods for experimental design such as A/B testing, observational study design, and mathematical methods (mainly statistics) for making valid inferences. We also have simple general 'models' that can be applied to a wide range of situations and may yield insights like a ‘Swiss army knife’ of things to try, such as when testing new products and services in a market. But, unfortunately, that is not enough to motivate efficient progress of development of products, services and organisation. What data science for business also needs is a frame of reference that allows us to select hypotheses with more chance that the hypothesis is correct. This is because, like physics, a frame of reference would encode vital information about the context that applies, and what information is relevant and what is not, to our business problem. We could then use the frame of reference to better predict what changes will be beneficial to our product or our business and choose better experiments and analysis. We would then avoid doing experiments that are doomed to failure, or are only successful sporadically, or involve misrepresentations of the problem.
The limits of data science without a frame of reference
It is not enough to just pick and choose from a range of models, or even to be pluralistic in your model. You also need theory. A mature science is not only about 'data' or models but leverages theory with real insight. Scientific theory is of a higher level of information content than ordinary ideas and hypotheses, (despite the word ‘theory’ it is information that can be used with high confidence). Decision theory, psychology, economic theory and social psychology are also not enough for the business context because they have a different context or focus than the viable organisation. To use theory in data science for business we need a theory that sets the frame of reference at an organisational level, not a psychological or economic level. This is because organisations involve both agents, human and technological, technical systems, culture, policies, rules, incentives and systems of governance. Models and strategies etc, can then be chosen appropriately given an organisational context because an organisational theory sets the frame of reference which validates the potential use of a viable model or strategy in that overall context, not just a part of it.
In management science, as in cognitive science and sociology, as a substitute for a mature theory there is often the awareness that 'context is everything', but there is no clear definition of a frame of reference that allows you to define the relevant context. It remains undefined, and hence management science is not a mature scientific discipline, in contrast to physics.
The same is actually true of economics, which in lieu of a context-sensitive theory, uses axioms to simplify and enable an analytical approach, even though it is known that the real empirical context that the theory is applied to, often invalidates the axioms of the analysis. Also, the excessive use of ceteris paribus outside of any truly context-sensitive theory is unscientific. Ceteris paribus is the idea that some model is only true assuming ‘all other things being equal’, but without a way to check if this is actually the case, or what to do if it is not the case.
The promise of biology for organisational theory
However, biology is getting to become a mature discipline, the discovery of DNA, (the double helix) and various other discoveries before and since, plus the development of computational and other analytics methods in the domain of biology means that the theory is becoming more and more successfully developed. As a result, biological theory can explain the relevant context for a wide range of scientific questions about the complex organisation of viable organisms, in general. This is not referring only to ecology or the study of groups of large animals, but also, or especially, knowledge of the evolution of microbiological systems, such as bacteria, and multi-cellular life. By leveraging biological theory in the round, we can identify a context sensitive frame of reference for applying theories and sub-theories to organisational contexts. That frame of reference would be informed by a theory that roughly tells us what matters and why in any given context, and how to identify that relevant context with high confidence, at the organisational level.
Another analogy to understand why we need biology theory in business
To summarise, a huge amount of scientific research to do with adaptation, management of uncertainty and risk of viable organisations, is and has already been done for free (or rather, with your government's tax revenue) by researchers in biology, researching the same things in organisms which are controllable models of organisations tackling similar problems of viable development to businesses.
What you lose in precision by looking for information from the biology domain, you more than gain in scientific control of experiments, generality, etc., within that biological domain, leading to high quality, accurate, theories.
Using biological theory in business is a way to access a huge amount of academic research and theory development in a domain that is like businesses, but easier to do research within the domain scientifically. This is because the access to information you gain in biology is far more than the information you can gain over organisations as the kinds of experiments you can do are far superior to management science. While human organisations develop over decades and are hard or impossible to do controlled experiments on, multi-cellular and single celled organisms allow for a huge amount of experimental control and the research of a huge range of environmental challenges to development, adaptation, growth, the management of risk, etc.. As a result, our knowledge of biological systems is far in advance of our knowledge of human organisations.
Mouse models as an analogy
An analogy of learning from other domains from within the biology domain is how medical researchers use what are termed mouse models of human diseases to do research into promising drugs and learn more about the likely effects of treatments on humans. This is because mice are similar enough to humans for the information to be useful. One might ask: Why don’t they just study those diseases in humans? Well, that would give you more precise information, but what the researchers lose in precision by studying mice instead, they more than make up for because studying mice are easier to learn about scientifically than humans. This is because scientists can design better experiments, which give them more high quality, accurate, information, and mice are actually still quite similar to humans.
Mouse models explained
Mouse models are genetically engineered mice which are bred specifically to be susceptible to diseases similar to the humans diseases we care about. Examples are mouse models of high stroke risk or diabetes or Alzheimer’s. The engineered disease in the mouse is not quite the same because mouse genes are different to human genes, but also still quite similar. So despite the differences between mice and humans, scientists learn more about the human disease in the end, by studying a similar disease in mice. This is because they have more control over the experiment if the subjects are mice. It's easier to gain high quality scientific information about mice diseases, because mouse lifespans are short enough, for instance, that you can see the effects of mouse diets and drug treatments over the whole lifetime of the mouse and repeat this experiment many times. For a human that's not a practical experiment, so that information is not available with high reliability.
Viable organisations can be studied in biology
The same is true of the study of viable organisations. What we lose in precision by studying viable organisations of organisms instead of human organisation, we more than gain in terms of the ability to generate high quality empirical scientific data of real value which allows us to develop high quality theories to explain our observations.
It is arguable that not studying methods of analysis that can be applied to both biological organisation and economics organisation is like ignoring the potential to learn from other animals about human disease. Just because they are not the same, it doesn’t mean the information isn’t valuable, or even the most important source of high quality scientific information about humans. This is because the differences are also a huge advantage in terms of generating empirical data and developing scientific theory of viable organisations. Hence, we should all be studying ‘mice’ and many other simpler organisms such as bacteria, to learn high quality information about the science of viable business organisations, just as we learn from mice higher quality information about human diseases.
Life history theory as a frame of reference imported from biology
One of the main frameworks for defining the relevant frame of reference to understand viable organisation of organisms is the development over several decades of research, of biology’s life history theory. It functions as a way of grouping sub-theories of viable organisms together. Life history theory identifies high quality viable strategies based on the relevant aspects of the environmental context, a relationship which is explicitly defined in terms of the theory of life history of an organism.
It is not possible to come up with a law in business like the Law of Conservation of Energy as the environment and the problem is too complex. It is, however, possible to come up with ‘20 good questions’ for people developing products and services in business, based on the structure of life history theory, as a high quality theory from biology.
These are the ‘20 good questions’ to ask in a given business context or in an organism’s context to identify what strategies an organisation should adopt to be viable in that challenging business context. This is ‘good’ versus many other questions which wouldn't yield as useful and productive answers, on the average. This addresses the business problem of knowing what experiments to do, or questions to ask, to give the best chance to improve a product, service, or the organisation of a team, department or business. So let’s explain how 20 good questions can form a frame of reference when the questions to ask are informed by biology's life history theory.
The concept of ‘20 Good Questions’ as a frame of reference
The game 20 Questions is a game that is played where one person thinks of a thing, person or animal and the other players can take turns asking questions of the form that allow only a ‘yes’ or ‘no’ reply. They only have 19 questions before they have to guess what it is that the question answerer is thinking of. This game is best played by asking questions that remove as close to 50% of the possible answers at each step. In the same way, asking the right questions of your business problem, to get close as possible to the best answer about what your strategy for your product or service development shoud look like, relies on a theory about which questions about the business context have the most information content.
The 20 Questions game is also the basis of patented algorithms that are used to guide missile systems because they are about the principle of navigating in uncertain, noisy environments and an example of the application of information theory, to guess the better questions to ask of the data coming into the missile guidance system.
There are thousands of questions you could ask: Some of these questions are like asking (with no prior knowledge):
Q:What the is the current favourite hypothesis of the executive, (who has only a hunch but no reliable information)?
A better question is asking :
Q:Which product features will be most successful for the target market we are launching in?
Hence, some questions get us closer to the viable product, and some questions are a waste of time, just like the game 20 Questions.
Turing Meta's hypothesis for businesses, is that the frame of reference, is just the 20 good questions to ask of a business to identify the viable strategy for that business problem and how these questions are related, so that given a certain answer to one question, which identifies the business context, the next question to ask given that business context is then question x, and so on.
The second hypothesis is that these 20 good questions and how they relate to each other is very similar to the 20 good questions to ask biological systems to identify the viable strategy for an organism in a given kind of environment:
First we ask the good questions about the environment, which is the most relevant context and challenge we have to deal with to produce a viable organism in that environment, then we ask the relevant questions to check what kinds of strategy, with options predicted by biology’s life history theory we should choose to investigate further, given the state of the environment we need to survive and thrive in.
So the structure of these 20 good questions as a frame of reference from biology mainly hangs off of life history theory. An example of one such question is to ask is:
Q: What is the level of uncertainty in your business environment?
When you know that, you then know whether you should seek then investigate whether an r or K strategy as a business or for a product is more likely to be viable, which I explain in this article.
Testing Phase Example
If I am responsible for a specific life stage or phase for a product or service, such as the product testing phase, e.g., for software development and I want to optimise for a successful testing phase, 20 good questions based on life history theory, to ask include:
Q: In this environment, what is the cost of not getting testing 100% right on the next life stage of the product and for the phases or life stages up to maturity of the product?
Q: What is the predicted lifespan of the product?
Q: What is the rate of growth we need to be competitive?
The second and third questions might seem odd. But biology shows us clearly, that in the case of the testing phase of a product, the analogy for organisms is the cost of preventing errors versus the risk of survivable errors in the next generation or life stage. Depending on the relative risk, different strategies are best investigated and change the reproductive rate (or rate of growth) of the organism/population.
In biology we see different levels of effort to avoid genetic mutations. High effort is expensive in terms of energy, so that there are differences depending on how reliable the information transfer has to be versus the value of a higher reproductive rate, which is what you can do instead, when you save energy by checking less for errors.
We know that some long lived and large animals like elephants and whales invest far more in avoiding errors when copying DNA to replicate their cells because of their size and long lifespans which means that even small errors will lead to a high cost over time.
Smaller shorter-lived animals and organisms can afford to invest less in copying errors, which might also have a positive effect, as this can be strategy to boost feedback via natural selection for new organisms and raise the reproductive rate. A predicted high growth rate can be met for shorter lifespan animals and organisms by lowering the quality of error detection.
These questions won't always be necessary, or the best questions to ask in a business, but it is predicted by life history theory that strategies of successful organisations are sensitive to these questions about the environment. The questions therefore have a high probability of useful information because of the similarity to challenges of viability for biological organisms.
Understanding the context sensitive predictions of life history theory, we can predict when this question is a good question to ask and what detailed questions need to be asked to get to a viable product and service in which all hoops or life stages are successfully negotiated in a viable way.
Common sense, but more so
Leveraging biological science in business by using the correct analogies, is often like using common sense, but as the mathematician Jordan Ellenberg says, common sense with a bionic exoskeleton. Science makes that common sense so much more powerful. You are standing on the shoulders of decades of biological research, and accessing that careful thought and high quality experimentation through carefully worked out theories that have relevance across domains and have been thoroughly tested. And sometimes the next good question to ask is actually counter-intuitive. So using theory to motivate questions allows you to develop applications of the theory with confidence, as long as you have thought carefully about what to focus on and the structure of the analogies to businesses.
Turing Meta’s research is not the only research into economics leveraging biological research and vice versa:
See evolutionary game theory applied to economics.
See tropic theory applied to macroeconomics
See trading strategies between funghi and plants applied from economics to biology
See computational approaches to economics and biology which seek cross disciplinary principles, such as using information theory, developing ideas from thermodynamics and entropy etc.
In conclusion, making the analogy between biology and organisational theory is a kind of scientific 'meta-analogy'. It is a template for constructing useful scientific analogies between biology and organisations and leveraging high quality theory for your business problems.
To understand more about how to apply biological theory as a high quality frame of reference for your business problems see other articles on www.turingmeta.org, subscribe or contact Turing Meta directly to discuss training opportunities and bespoke analysis.