Business Science and its Enemies

#FollowingTheScience As politicians around the world struggle to contain and master the Covid-19 pandemic, the complex role of science in guiding decision and policy has been brought into view. Not only the potential tension between science and policy, but also the tension between different branches of science. (For example, medical science versus behavioural science.)

In this post, I want to look at the role of science in guiding business decisions and policies. Professor Donoho traces the idea of data science back to a paper by John Tukey in the early 1960s, and the idea of management science, which Stafford Beer described as the business use of operations research is at least as old as that. More recently, people have started talking about business science. These sciences are all described as interdisciplinary.

Operations research itself is even older. It was originally established during the second world war as a multi-disciplinary exercise, perhaps similar to what is now being called business science, but it lost its way in the years after the war and was eventually reduced to a set of computer programming techniques with no real impact on organization and culture. 

In a recent webinar on Business Science, Steve Fox asked what business science enabled leaders to do better, and identified three main areas. 

Firstly system-related - to anticipate requirements and resources, identify issues, including risk and compliance issues, and fix problems. 

And secondly people-related - to tell the story, influence stakeholders and negotiate improvements. Focusing on message and communications to the various audiences we need to influence is a key part of business science. 

And thirdly, thinking-related. When business science is applied correctly, it changes the way we think. 

I agree with these three, but I'd like to add a fourth: organizational learning and agility. This is an essential component of my approach to organizational intelligence, which is based on the premise that serious business challenges require a combination of human/social intelligence and machine intelligence.

Steve Fox also stated that the biggest obstacles to creating data-driven business aren't technical; they're cultural and behavioural. So in this post, I also want to look at some of the obstacles of following the science in the context of business and organizational management.

• Poor Data - Inadequate Measurement and Testing - Ignoring Critical Signals
• Too Much Data - Overreliance on Technology - Abdication
• Silo Culture - Someone Else's Problem
  
• Linear Thinking - Denial of Complexity
  
• Premature Attempts to Eliminate Uncertainty
• Quantity becomes Quality

After I had initially drawn up this list, I went back to Tukey's original paper and found many of them clearly stated in there. 

---

Poor Data - Inadequate Measurement and Testing - Ignoring Critical Signals

Empirical science relies heavily on a combination of observation, experiment and measurement. 

Too Much Data - Overreliance on Technology - Abdication

Tukey: Danger only comes from mathematical optimizing when the results are taken too seriously.

Adrian Chiles reminds us that all the data in the world is no use to you if don’t know what to do with it. He quotes Aron F. Sørensen (via Chris Baraniuk) Maybe today there’s a bit of a fixation on instruments.

And in many situations, people overrely on algorithms. For example, judges relying on algorithms to decide probation or sentencing, without applying any of their own judgement or common sense. If a judge doesn't bother doing any actual judging, and lets the algorithm do all the work, what exactly are we paying them for?

Linear Thinking - Denial of Complexity

Tukey: If it were generally understood that the great virtue of neatness was that it made it easier to make things complex again, there would be little to say against a desire for neatness.

One of the best-known examples of linear thinking was a false theory about the vulnerabilities of aircraft during the second world war, based on the location of holes in planes that returned to base. People assumed that the vulnerabilities were where the holes were, and this led to efforts to reinforce planes at those points.

Non-linear thinking turns this theory on its head. If a plane makes it back to base with a hole at a particular location, this should be taken as evidence that the plane was NOT vulnerable at that point. What you really want to know is the location of the holes in the planes that did NOT make it back to base.

In 1979, C West Churchman wrote a book called The Systems Approach and its Enemies, about how people and organizations resist the ways of thinking that Churchman and others were championing. Among other things, he noted the way people often preferred to latch onto simplistic one-dimensional/linear solutions rather than thinking holistically.

---

Chris Baraniuk, Why it’s not surprising that ship collisions still happen (BBC News 22nd August 2017)

Christa Case Bryant and Story Hinckley, In a polarized world, what does follow the science mean? (Christian Science Monitor, 12 August 2020)

Adrian Chiles, In a data-obsessed world, the power of observation must not be forgotten (The Guardian, 5 November 2020)

C West Churchman, The Systems Approach and its Enemies (1979)

David Donoho, 50 years of Data Science (18 September 2015)

John Dupré, Following the science in the COVID-19 pandemic (Nuffield Council of Bioethics, 29 April 2020)

Faye Flam, Follow the Science Isn’t a Covid-19 Strategy: Policy makers can follow the same facts to different conclusions (Bloomberg, 10 September 2020)

Steve Fox, A better framework is needed: From Data Science to Business Science (Consider.Biz, 17 September 2020) via YouTube

Matt Mathers, Ministers using following the science defence to justify decision-making during pandemic, says Prof Brian Cox (Independent, 19 May 2020) 

Megan Rosen, Fighting the COVID-19 Pandemic Through Testing (Howard Hughes Medical Institute, 18 June 2020)

John Tukey, The future of data analysis (Annals of Mathematical Statistics, 33:1, 1962)

Wikipedia: Data Science, Management Science 

Related posts: Enemies of Intelligence (May 2010), Changing how we think (May 2010), Data-Driven Reasoning - COVID (April 2022)

Enterprise Architecture as Science?

It is common to describe Enterprise Architecture as a science. Here are a few examples.

• ... the evolving science of Enterprise Architecture ... (Center for Enterprise Architecture, Penn State University)

• We see enterprise architecture (EA) as a scientific sub-discipline both of computer science and business management. The twice mentioned word “science” here emphasizes our certainty that EA is an exact discipline able to produce precise approaches and solutions. Wolf Rivkin, Enterprise Architecture and the Elegant Enterprise (Architecture and Governance 5-3)

• At iCMG, we have proven beyond doubt that “Enterprise Architecture is a scientific basis for creating, maintaining and running an Enterprise". Sunil Dutt Jha, Biggest myth - “Enterprise Architecture is a discipline aimed at creating models” (Architecture World, January 2013)

A few years ago, I discussed this question with @RSessions

• Is Enterprise Architecture a Science? (July 2009)
• Is Enterprise Architecture a Science (2)? (November 2009)

Roger is one of the few people I know who is seriously committed to empirical investigation of EA. I believe he shares my view that much EA falls woefully short of anything like scientific method. To my eye, many knowledge-claims within the EA world look more like religion or mediaeval scholastic philosophy than empirically verifiable science.

But why does it matter anyway? Why would people be so keen to claim EA as a science? Here is what Foucault had to say to those who wished to claim Marxism (or psychoanalysis) as a science.

"When I see you trying to prove that Marxism is a science, to tell the truth, I do not really see you trying to demonstrate once and for all that Marxism has a rational structure and that its propositions are therefore the products of verification procedures. I see you, first and foremost, doing something different. I see you connecting the Marxist discourse, and I see you assigning to those who speak that discourse the power-effects that the West has, ever since the Middle Ages, ascribed to a science and reserved for those who speak a scientific discourse." Michel Foucault, Society Must be Defended: Lectures at the Collège de France, 1975-76 (English translation by David Macey, 2003)

In other words, claiming EA as a science is not about the rational basis for its knowledge-claims but about its authority, or what Foucault (in David Macey's translation) calls Power-Effects. Thus instead of claiming EA as a science, one might follow Gartner in claiming EA as a discipline.

Enterprise architecture (EA) is a discipline for proactively and holistically leading enterprise responses to disruptive forces by identifying and analyzing the execution of change toward desired business vision and outcomes. EA delivers value by presenting business and IT leaders with signature-ready recommendations for adjusting policies and projects to achieve target business outcomes that capitalize on relevant business disruptions. EA is used to steer decision making toward the evolution of the future state architecture. (Gartner website, retrieved 17 August 2013)

Foucault characterizes a discipline in terms of the selection, normalization, hierarchicalization and centralization of knowledge. We can surely recognize these processes in the formation and maintenance of EA frameworks such as TOGAF and PEAF, as well as various attempts to construct Bodies of Knowledge. Foucault notes that "the progress of reason" necessitates "the disciplinarization of polymorphous and heterogeneous knowledge". This might lead us to expect some institutional resistance to heterodox ideas, as well as the marginalization of "amateur scholars".

Foucault is interested in ways that people and organizations can respond to disruptive forces large and small, from "great radical ruptures, massive binary divisions" to "mobile and transitory points of resistance, producing cleavages in a society that shift about, fracturing unities and effecting regroupings".

Just as the network of power relations ends by forming a dense web that passes through apparatuses and institutions, without being exactly localized in them, so too the swarm of points of resistance traverses social stratifications and individual unities. And it is doubtless the strategic codification of these points of resistance that makes a revolution possible. [Michel Foucault, History of Sexuality Vol 1.]

Gartner's notion of EA-as-discipline seems quite consistent with this. It is focused on mobilizing the response to disruptive forces (for which Gartner uses the rather strange word "nexus"). EA gains its power from a kind of strategic codification (or discursive practice), allowing the enterprise to "harness the nexus", thereby "revolutionizing business and society, disrupting old business models and creating new leaders". (Gartner website, retrieved 17 August 2013)

---

Update

@tetradian commented on the dangers of spurious 'authority' - 'spurious' in sense of claiming an aura of 'authority' when there's none to be had (b/c it isn't 'science' anyway)

I agree that claims of scientific status or method in the EA world are generally spurious. But there are other ways of asserting authority. For me, the key question is why (and on what grounds) should anyone trust the pronouncements of EA. It is not just about danger versus safety, but about authority versus authenticity.

From information architecture to evidence-based practice

@bengoldacre has produced a report for the UK Department for Education, suggesting some lessons that education can learn from medicine, and calling for a coherent “information architecture” that supports evidence based practice. Dr Goldacre notes that in the highest performing education systems, such as Singapore, “it is almost impossible to rise up the career ladder of teaching, without also doing some work on research in education.”

Here are some of his key recommendations. Clearly these recommendations would be relevant to many other corporate environments, especially those where there is strong demand for innovation, performance and value-for-money.

• a simple infrastructure that supports evidence-based practice
• teachers should be empowered to participate in research
• the results of research should be disseminated more efficiently
• resources on research should be available to teachers, enabling them to be critical and thoughtful consumers of evidence
• barriers between teachers and researchers should be removed
• teachers should be driving the research agenda, by identifying questions that need to be answered.

Clearly it is not enough merely to create an information architecture or knowledge infrastructure. The challenge is to make sure they are aligned with an inquiring culture.

to be continued ...

---

Ben Goldacre, Teachers! What would evidence based practice look like? (Bad Science, March 2013)

Arguing with Mendeleev

@JohnZachman insists that his classification scheme is fixed—it is not negotiable. Comparing his Zachman Framework with the periodic table originally developed by Dmitri Mendeleev, he says, "You can't argue with Mendeleev that he forgot a column in the periodic table".

Well, actually, you can. If you look at the Wikipedia article on the Periodic Table, you can see the difference between Mendeleev's original version and the modern version. Chemists nowadays use a periodic table with 18 columns. As Wikipedia states, "Mendeleev's periodic table has since been expanded and refined with the discovery or synthesis of further new elements and the development of new theoretical models to explain chemical behavior."

That's what makes chemistry a true science - the fact that the periodic table is open to this kind of revision in the light of experimental discovery and improved theory. If the same isn't true for the Zachman Framework, then it can hardly claim to be a proper science.

Some observers have noted that early versions of the Zachman Framework had fewer columns, and see this as a sign that the number of columns may be variable and open to discovery. They also interpret the word "extending" in Zachman's 1992 paper (with John Sowa) as an acknowledgement that the framework has evolved. But the Zachmanites reject this: they say that the six columns have always existed, it was just that the early presentations didn't mention them all. "Humanity for the last 7,000 years has been able to work with what, how, who, where, when, and why." (This sounds like a Just-So-Story - "How the Enterprise Architect Got His Toolset".) Questions that require more than one word in the English language (such as How Much and For Whom) can be discounted. (Graeme Simsion made this point in 2004, and this may have been what prompted Scott Ambler to add a cost column to his extended framework.)
 
Mr Zachman has a degree in chemistry, so he ought to understand what makes the periodic table different from his own framework. However, some of his followers are less cautious in their claims. I found an article by one Sunil Dutt Jha, whose "proof" of the scientific nature of EA seemed to rely on two key facts (1) that Mendeleev transformed alchemy into chemistry by creating the periodic table, and (2) that the Zachman framework looks a bit like the periodic table, therefore (3) EA must be a science too.


An earlier version of this comment was posted on Linked-In Is it true to say that “Enterprise Architecture” is a scientific basis for creating, maintaining and running an Enterprise?

---

Scott Ambler, Enterprise Agile: Extending the Zachman Framework (undated)

Philip Boxer, Modelling Structure-Determining Processes (19 December 2006)

Sunil Dutt Jha, Biggest myth – “Enterprise Architecture is a discipline aimed at creating models” (January 2013)

Graeme Simsion, What's wrong with the Zachman framework? (TDAN, January 2005)

John Sowa and John Zachman, Extending and formalizing the framework for information systems architecture (IBM Systems Journal Vol 31 No 3, 1992)

Ivo Velitchkov, Frameworks and Rigour (3 March 2013)

Alan Wall, Pattern Recognition and the Periodic Table (March 2013)

John P. Zachman, The Zachman Framework Evolution (2009-2011)

Erecting the Framework (Feb 2004) - John Zachman discussing his Zachman Framework for Enterprise Architecture in an interview with Dan Ruby



Related posts and presentations

For Whom (November 2006), The Kipling-Zachman lens (June 2009), Deconstructing the Grammar of Business (June 2009), Satiable curtiosity (September 2009), Evolving the Enterprise Architecture Body of Knowledge (October 2012), Enterprise Architecture as Science (August 2013), What is a Framework (February 2019)

eBook: Towards Next Generation Enterprise Architecture

Updated 04 May 2019

Alignment - Science or Pseudoscience?

@RSessions claims that "Simplification occurs when the IT partitions align with the business partitions". But what does alignment mean?

Roger defines alignment as "a measure of how well two patterns overlay on top of each other". But this definition only works between two patterns that already occupy the same space.

For example, I think I know what it means to say that the furniture is aligned with the walls of the room, because I can measure this fact with geometrical instruments (ruler, set square), but if someone says that the colour of the furniture is aligned with the sensibilities of the owner, I can only make sense of this as a vague metaphor rather than a precise measurable fact.

The fallacy of astrology does not lie in the detailed analysis of alignments between the planets and stars - which after all occupy the same astronomical space - but in the notion that the movements of the planets against the stars are somehow aligned with the patterns of human affairs on earth. (And the reason this counts as pseudo-science is that the detailed correlation between sky and earth relies on an interpretation by an astrologer.)

When people talk about business-IT alignment, I can only make sense of this as a metaphor. I am not aware of a robust and meaningful formalization that would permit both business and IT to occupy the same geometrical space, and I can't see the point of inventing one.

All we need is a mapping between two patterns occupying different spaces. An alignment is a special kind of mapping within a fixed geometrical space. If we can't define a fixed geometrical space, then I regard the concept of "alignment" is a misleading metaphor, introducing unnecessary complication. I prefer the general concept of mapping to the false metaphor of alignment.

@aleksb6 objects that my position is "true iff two patterns to map. reality is a M2M relationship, so 'alignment' is not a complication, it's a necessity!" He continues "btw, isn't mapping two patterns just another instance of point-to-point thinking? I thought we wanted to discourage that!"

If alignment doesn't make sense between two things, I can't see that it makes any more sense between three or more things. The desired outcome is a set of structure-preserving mappings between as many things as we need to coordinate. That doesn't mean we can or should design each mapping separately. In all but the most trivial situations, however skilfully we decompose a large problem into subproblems, there is always some coordination (juggling) left to do.

Here's the bottom line. If assertions about business-IT alignment are to mean anything at all, then you have to have a way of looking at a lump of business and a lump of IT and say whether they are aligned or not, and if so how much.

Of course, if you believe you have a modelling language that can express both business and IT, then you might think all you have to do to find out if business and IT are aligned is to compare two models. But this turns out to a circular procedure, because the modelling languages are themselves justified by the claim that they promote business-IT alignment, so we cannot use the modelling languages themselves to prove that business-IT alignment has been achieved. There has to be some point of reference outside the modelling languages.

People keep telling me "alignment" is important, but they can only define it as a woolly and subjective metaphor. So if we stop worrying about "alignment", and talk instead about the various multi-dimensional mappings between a complex system of systems and a complex set of business requirements, we can concentrate on what is objectively important.

And leave the concept of "alignment" for the astrologers. All together now ...

When the Moon is in the 7th house and Jupiter aligns with Mars
Then peace will guide the planets and love will steer the stars.

Is Enterprise Architecture a Science? Part 2

@RSessions was in London this week, so I sat down with him to continue our previous discussion Is Enterprise Architecture a Science?

The first question to address is - which enterprise architecture are we talking about? I think we both agree that there are some activities within the EA world that look more like religion or mediaeval scholastic philosophy than empirically verifiable science.

For example, in his post What's Right with the Zachman Framework, Grant Czerepak states that "the architectural metaphor conceals what the six perspectives are actually about: Entities, Relationships, Attributes, Constraints, Definitions and Manipulations". And referring to the Kipling-Zachman lens, Grant claims that "the interrogatives have a foundation that goes back over three thousand years across every human culture". In a separate post, he says It's not Aristotle's fault, it's your fault. See my post Arguing with Mendeleev (March 2013).

Lots of EA frameworks are essentially abstract classification schemes that start from an abstract ontological argument ("obviously all businesses are made of objects" or "obviously all processes are made up of nouns and verbs") and make assertions that are not amenable to empirical verification.

Roger's SIP methodology is at least based on an empirically testable (and quantified) hypothesis. That a system of systems with such-and-such measurable structural qualities (in terms of Roger's definition of complexity) will have such-and-such predictable costs. So this provides the basis for a scientifically-grounded engineering practice. I think SIP methodology has a reasonable claim to be scientifically grounded: it can be evaluated not just on whether its prescriptions are practical, cost-effective and useful, but also whether its predictions are true. (Incidentally, I think it would be interesting to compare SIP with Christopher Alexander's early book Notes on the Synthesis of Form. This contains detailed and mathematically grounded work on architectural complexity: although the software gurus who developed structured methods in the 1970s were aware of Alexander's book, they left out most of the difficult detail.)


I think there is a further step before EA could ever dream of becoming a fully empirical science, and this would involve large-scale collection and analysis of empirical data, so that there would be a closed loop between theory and practice, connecting structure and value. In order to achieve this, we should need the active participation of some of the more powerful players in the EA game - the large consultancies and above all the key government agencies that govern IT expenditure. (You know who you are.) At the moment, there is little sign that these organizations are seriously interested in any game-changing innovation. (Roger and I should be delighted to talk to representatives of these organizations, please contact us.)

Is Enterprise Architecture a Science?

asks @RSessions (Roger Sessions)

I have long argued that the answer is no. What passes for knowledge in enterprise architecture is largely a combination of anecdote and received opinion. Intellectual effort is devoted to hierarchical forms and elaborate classification schemas, based on abstract reason rather than empirical measurement; this kind of work looks more like mediaeval scholastic philosophy than modern science.

In comparison, the followers of Christopher Alexander seem like nineteenth century gentleman scientists, carefully collecting specimens from which they try to infer useful principles and patterns. Alexander's four-volume masterpiece on the Nature of Order is a brilliant and fascinating work, which I think every enterprise architect should read (in their spare time), but it's not exactly suitable as an everyday handbook.

In order for enterprise architecture to qualify as a science, it has to follow scientific method. Now I am pretty broadminded about what counts as scientific method, but it's more than mere predictability. (Card games are predictable, but that doesn't make cribbage a science. Roger says if card games were predictable, he'd be a rich man. But if card games were not predictable, the casinos would go broke.)

Anyway, you can get predictability from art. I saw Dave Crosby on a documentary once, criticizing the fact that the Eagles' concerts were note-for-note predictable, CSNY preferring a slightly looser style in response to each audience. So what if EA's an art, asks @HotFusionMan (Al Chou).

Peirce understood the limitation of scientific method, holding "that slow and stumbling ratiocination can be dangerously inferior to instinct, sentiment, and tradition in practical matters" (Wikipedia).

So does it matter if (as I believe) Enterprise Architecture is not a science? The key question that raises is - what is the source and status of EA knowledge, and how can we ever resolve matters of opinion, except by obscure mediaeval argument?


Follow-up post Is Enterprise Architecture a Science (2)?