Managing Excerptations: Rationality

I started my blog to kick my non-writing habit. I've now got enough material to start blocking out a book (hooray!). So while I'm doing that, I thought I'd start a small series looking at some of the books that have influenced me.

"Managing expectations" is one of my least favourite managerialisms. I cringe whenever I hear it. Expecting is about hope, which is the only thing that keeps many professionals going. We take what we can from life and do our best with it.

So this series is about some of the excerpts that caught my eye over the last six months. They're part of what gives me hope about the profession of management.

I came across Keith Stanovich's 2010 book 'Decision Making and Rationality in the Modern World' when I found myself in a mild panic after realizing that I had no clear sense of what 'being rational' really meant.

I discovered that my sense of being rational was complicated. It had grown over time into a misshapen lump of an idea and I couldn't get clarity about what it really was.

This was a big deal for me because I spend a lot of time thinking about the impact of analytics on large organizations. Rationality is a fundamental concept for effective enterprise analytics investments. Everyone likes to be 'evidence' driven, but without a clear idea of what rationality is, anyone can (and does) simply pick the evidence that suits their needs.

Stanovich,

"To think rationally means taking the appropriate action given one's goals and beliefs, and holding beliefs that are commensurate with available evidence - but it also means adopting appropriate goals in the first place. Instrumental rationality covers the first of these and epistemic rationality covers the second, but the third factor introduces a new issue. That issue is the distinction between a thin and broad conception of rationality."

This excerpt sums up very well the complicated view of rationality. Instrumental rationality is about 'what to do' and epistemic rationality is about 'what is true'. The business world tends to focus on instrumental rationality, following encouragement by an influential stream of business thinkers from Frederick Taylor to Milton Friedman.

Epistemic rationality is what people talk about when they say 'evidence driven' and is the domain of the analytics industry. Whenever we read a study ("x% of business executives eat bananas for breakfast") we're in the world of epistemic rationality. Some business professionals are coming to seriously examine this world and we should all be thankful for it.

But it is the idea of thin and broad conception of rationality that we (finally) start asking questions of morality. Anyone who's read my work knows that this issue of business morality is one of my principal questions about the way management professionals conduct themselves.

What is this thing called 'thin rationality'? Why is it something that responsible management professionals need to beware of? Stanovich explains,

"The models discussed in the bulk of this book have been thin theories of rationality. They are means-ends models: they assess whether people are using efficient means to reach their previously existing desires and goals. Thin theories take beliefs and desires as given and simply ask whether people choose the right action given their beliefs and desires. Specifically, the content of desires is not evaluated."

"If all we aspire to as humans is a thin theory of instrumental rationality, then human rationality is indeed just like chimpanzee rationality in that rational choice would be evaluated in the same way in both cases. We would ask whether the sets of actions adhered to the axioms of choice."

"But there is every reason to believe that most people do not want to stop with a thin theory. Most people take a stance toward the choices they make and the goals that they pursue. Both the choices and the goals are often evaluated by external criteria. The choices are evaluated in terms of the meaning they convey to the person making the choice and the goals are evaluated in terms of whether they are consistent with the values the persons holds."

"Our values are the main mechanism we use to initiate an evaluation of desires. Action/value inconsistency can signal the need to initiate normative criticism and evaluation of both the first-order desires and the values themselves. Values thus provide an impetus for the possible restructuring of the architecture of desires. They are what allow human rationality to be a broad rationality - one where the context of desires makes a difference - in contrast to the thin instrumental rationality characteristic of chimpanzees."

What I take from this is that it is absolutely critical for management professionals to place values at the center of all they do. If we do not, then we are essentially giving up one of the great benefits of humanity: the capacity to choose to do good.

Because machine learning has a pronounced advantage over humans in algorithmic decision making, we no longer have much of a future with routine work. If the next best thing is to exist at the same decision making level as chimpanzees then I for one want an upgrade out of business class.

The management professional of the future is characterized by her willingness to confront "some very difficult issues in philosophy and cognitive science". If we don't, then all we have is a choice between being a chimpanzee or a computer.

Image via Gratisography