Most approaches to culture change are ill informed. Cognitive science provides a better way.
Part 1 of 2.
These are heady times for those of us in the culture business — which is pretty much anyone who leads an organization or tries to help those that do. Organizational “culture” seems to explain everything, from Uber’s rapaciousness to Amazon’s intensity to GM’s ethical lapses to Enron’s greed. Startups now have VPs of “Culture”. The term is so firmly in the mainstream that athletes and sports managers use it to explain why, say, basketball’s Golden State Warriors are so good, or why European football’s Arsenal are so perpetually mediocre.
Wanting to improve your organization’s inclusion, creativity, performance, and overall humanity is laudable. Unfortunately for us cognitive anthropologists, the culture fad is like Dr. Frankenstein’s monster escaped from the lab to wreak havoc on the countryside. A well-intended and promising invention gone alarmingly awry. And most of us in the lab are at least partially to blame.
This piece explains why, and what can be done. I’ll begin by illustrating the problem through a short thought experiment to show how most of our assumptions about culture are ill informed. In Part II I will introduce the new science of culture and show how it might be leveraged.
The Mainstream Way
Imagine you are the CEO of a medium-sized, global industrial company (say 15,000 employees, 75 countries), who wants to change your company’s culture. Let’s say you want to do so because, well, your world is changing: product and service digitization, the Internet of Things, changing customer demand, the rise of emerging markets, and on. You believe your organization needs to move faster, be more innovative and collaborative in order to avoid being disrupted and rendered irrelevant in this new, rapidly changing post-industrial world. You believe a change to your culture is the key to avoiding this fate.
Where would you begin?
You might begin by impressing upon your direct reports and senior managers that they need to be more innovative and collaborative. You might say something like, “Culture change starts at the top.” Some of your colleagues might wonder what exactly you mean, to which you would say, “You set the tone. If you are more this way the rest of the organization will be more this way too.”
You might invest in internal marketing. This could mean a clever branding campaign with imagery and tag lines such as“Connect” or “Go Big” or “Make Something Happen!” to fill posters in conference rooms, tent cards in cafeterias, and new hire collateral. These themes become the centerpieces of your town halls as you evangelize the benefits of upping the collective innovation and collaboration quotient of your company.
Depending on the size of your budget, you might create training that teaches people how to behave in more collaborative and innovative ways. A sizeable chunk of your workforce will be required to take such classes, with everyone receiving a tchotchke reminding what innovation and collaboration really mean.
Rightfully you will be concerned these efforts are making a difference. So you will ensure the next employee engagement survey measures innovation and collaboration. In fact, a year passes and you are pleased to see the relevant survey questions have improved by a couple of percentage points from industry benchmarks.
So, after all this, have your changed your culture?
You might say you have, pointing to the survey results. To which I would ask, is your organization truly more collaborative and innovative? Or, might you be measuring the effects of priming by a marketing and training exercise? And is employee engagement the same as culture?
The scenario above, with minor variations, is the most commonly used approach and evidentiary standard for culture change in organizations today. Here’s the problem: most of the time it doesn’t work. The emerging cognitive science of culture helps explain why.
The 5 Myths of Culture Change
From the standpoint of research in the last 30 years on the relationship between the mind and culture emerging from cognitive anthropology, psychology, linguistics, sociology, cultural neuroscience and artificial intelligence — every aspect of this approach is flawed. The assumptions underpinning it are based either on outmoded theory, or no theory at all — when is the last time you read a popular book on culture based on actual theory or research? Remarkably, much of the mainstream approach is based on anecdote, wishful thinking, or trivialization: culture is operationalized in such a narrow or generalized way that “proving” something about it has little bearing on the lived reality of the organization. 
In other words, most of the ways in which we approach culture today are based on assumptions that upon closer examination reveal themselves simply to be myths with little or no foundation in modern research.
Myth # 1: Culture and culture change start at the top
This myth is so much part of the mainstream few of us have ever stopped and asked if it is supported by any independent evidence. It is based on an assumption of a linear and causal relationship between leadership and culture. That is, the leader shapes The Culture. There is another assumption embedded within this: that organizations have single, unified cultures.
Take the second one first.
The idea of a single, unified culture in an organization comes from literal interpretations of the classic ethnographies of the early 20th century on so-called “simple societies”, such as Mead’s work on Samoa or Benedict’s on the Zuni. While these groundbreaking ethnographers realized the concept of rigid cultural boundaries was of limited use and anthropologists abandoned it decades ago, many practitioners and management scholars persist with literal interpretations of this idea to this day, assuming any “society” such as an organization has a single culture.
Work in cognitive anthropology, sociology and linguistics shows us that cultural boundaries are highly permeable and fluid; people in organizations inhabit multiple pre-existing cultural systems (ethnic, regional, occupational, site or division-specific, etc.). To think of any sizeable organization having a single unifying culture is simply projection or wishful thinking. Cognitive science actually explains why we persist in believing in singular monolithic cultures, which I get to later on.
The assumption that leaders set culture has its roots in the Culture and Personality school of psychology from the first half of the 20th century. Management scholars extended the idea to organizations — some might say by a leap of faith — to suggest over time organizations can adopt the personalities of their leaders. There is undoubtedly a relationship between leadership personality and culture, but it is much more complex than that. For one, the relationship is mostly in the other direction: legacy cultures influence socialization. Simple relationships between CEO personality and culture have not been systematically observed in any organization beyond a small start up. And measuring such relationships is fraught with empirical problems: personality traits measure features of language, not culture; trait attribution systems vary across cultures; many CEO traits such as “achievement orientation” are so generalized in the CEO population that to suggest they occur as part of a particular culture is meaningless.
Yet these assumptions persist unchecked, for two reasons. First, since the 1970s leaders have been taught that organizational control through normative means (such as culture) is far more desirable than overt control (‘theory Y’ vs. ‘theory X management). Leaders tend to want to imprint their own legacies on their organizations. It is not surprising that this tendency is most prevalent in the U.S.: leader ‘self-enhancement’ is rooted in American cultural models of leadership benevolence and individual agency. We want to believe leaders shape culture because it is deeply consistent with our ideals of what leaders are supposed to do.
A more modern and scientifically sound way to think about the relationship between leadership and culture is through the lens of power. Leaders set agendas, allocate capital, marshal resources, and determine rewards. These actions influence organizational practices that in turn shape how people think and act.
Even in the most well publicized cases of charismatic leadership (ostensibly) shaping culture (for example, Ray Dalio at Bridgewater), on closer examination one can invariably identify actual organizational practices at the heart of the culture shaping. There is no straight line to be drawn between leadership and culture; at best, leaders may influence the conditions through which cultures take shape.
Tip: leaders would see a lot more ROI for their culture investment if they prioritized holding their organizations accountable for practices supportive of a desired future state than worrying about corporate values.
Myth # 2: Culture is shaped by language
The emphasis on internal marketing in our thought experiment, and on leaders “setting” culture through what they say is based on the assumption that language changes culture. We assume if we just get the words right, that will shape desired behavior — as if the problem is word choice. This is one reason why culture change is often approached as a problem for the corporate communications or PR team.
There is nothing wrong with clever catch phrases to help sell a vision or desired value. The problem is most companies stop there.
The emphasis on language reflects a many decades-old debate in cognitive science about whether language shapes, rather than reflects, how people think. But this debate has been largely resolved: cognitive linguists demonstrated convincingly we have many more thoughts, ideas and feelings than we can encode in language. Cultural knowledge is more than what can be conveyed through words alone. Therefore, to reduce culture to slogans or value statements is to mistake symptom for cause, or the mirror for the face. Language reflects some aspect of culture, yes, but it should not be assumed language is the same as culture. The prevailing view now is that language and culture in any society are parallel systems with overlaps and touch points that reflect, support, or at times serve as adaptations to each other. It is not the case that by simply espousing the culture we want we can change cognitive structure. In fact, based on how the brain processes novel versus conventional figurative language, it is likely that using typical business jargon as part of your change program will not be memorable at all.
Tip: Language matters, but less than you might think. Without supporting practices, your clever pronouncements or marketing will amount to very little.
Myth # 3: Culture can be trained
Myth # 4: Culture can be measured in opinion surveys
That culture can be learned through training, and that opinion surveys measure culture rely on the same problematic assumptions of boundaries and language. A corollary assumption is that culture is synonymous with values and normative precepts, and that these can be taught through training classes.
Again, wishful thinking.
The literature on values and culture is littered with more problems than can be described fully here. Among them: large organizations are made up of multiple value systems; there is a difference between personal and institutional values; values are idealizations, beliefs in how I think I should be; values tend not to be internalized in a population unless they are already well-socialized in that population; leaders in any society tend to have more values in common with each other than they do with the rest of the population; employees may be prone to saying (or responding on surveys) what they believe top management wants to hear; and people don’t always act according to what they believe.
Opinion surveys measure something: opinions. Attitudes and opinions are not culture. As with language, they might reflect some aspect of it, or be compensations for it. For example, a technology client of ours publicly values open communication and goodwill towards others. In our research there, however, we observe robust thinking and mastery of craft to be deeply held cultural models that orient much of what is considered “good” in the organization, and that in practice get in the way of “goodwill”: it is difficult to have goodwill when you don’t respect your colleague for his lack of craft. Openness and goodwill are publicly espoused, however, because they should be ideally important, especially given the pervasive orientations about what really “counts”.
Tip: “Training” a desired culture through norms and values will be highly problematic in any non-homogeneous organization (which is to say, almost all of them). At best, training might provide something to aim for, or better, be acknowledged as compensation for deeper cultural realities.
Myth # 5: Culture is a real “thing”
All of the above can be subsumed under the mother of all organizational culture assumptions: that culture is something tangible. From this all culture shaping proceeds. Managing culture is no different from managing any other asset: “target” what you want (e.g. more collaboration), then focus on “fixing it” (through marketing, training, etc. The words in quotes underscore the metaphor of a physical thing).
Making culture out to be a concrete object is due to the pragmatism of business: culture is, of course, a means to an end. It is also due to the fact the human mind tends to make sense of abstract things in terms of physical ones. Think of the way we characterize M&A as fighting, mating, or marriage — as if a merger literally is one of these things. We similarly talk and act as if culture is real and bounded, identifiable in language and in values, norms, and so forth. While its effects might be observed in these ways, this does not make it ontologically real in the same way, say, as a piece of machinery, or even money. For where, exactly, is that culture? Of what does it consist? What are its boundaries? Where does it come from?
Few executives and practitioners are able to answer these questions without resorting to the assumptions problematized above.
The Black Box of Culture Change
Which is one reason why so many culture interventions are, essentially, “black box”, consisting of sayings like be the change you want to be; change the game; keep it simple, etc., or else defaulting to loosely defined initiatives involving people (hiring, on-boarding, training) because, well, culture must have something to do with people (right?). Because we can’t readily “locate” culture, we are left trying to “solve” it through indirect means in hopes it will all work out.
We default to these kinds of interventions, well intended as they are, because they seem simple and easy to carry out. In truth, they are stabs in the dark, naïve attempts in the face of the fact we don’t really know much about the actual science of culture.
If culture engineering and change were really as simple as mainstream practitioners would have it, we would see much more empirical evidence of successful change, and a lot less academic debate. The culture literature would be filled with case after case of success. Recipes for change would be unambiguous. Culture would be more like accounting, or marketing: a body of knowledge on what works would be readily available.
Yet the opposite is the case. Companies have been interested in shaping or changing culture since the early 1980s, yet what we have to show for all this effort is that most managed culture programs fail to produce any sustained or measurable impact beyond anecdote (which usually comes from the very people espousing the culture change to begin with). Even when it is “proven” to have changed, it is usually because “culture” is operationalized as a single variable and shown, magically (through attitude or opinion surveys), to have moved in some direction.
There is a Better Way
Executives and practitioners are intuitively correct about culture’s fundamental importance and pervasiveness. The irony is that in trying to work with it, many dumb it down to single variables, simple recipes or hackneyed sayings based on little more than myth. Few would subordinate their marketing or financial strategy to such flimsy standards of evidence or research, but organizational culture work again and again proves the exception, unfortunately at great expense of time, energy, and resources.
Cultures do exhibit patterns and tendencies that are intuitively recognizable and exert a deep, if mostly hidden force on social systems. The key question is whether these patterns and tendencies can be engineered to produce predictable outcomes. The jury is still out. But modern science is moving us away from most of the current practices we see in organizations.
When what we are learning about culture from the cognitive sciences trickles down to the mainstream, we will witness a radical transformation in the practice of organizational effectiveness and change. That this knowledge is not (yet) widespread is partly because anthropology, the one discipline that puts culture at the center of its intellectual agenda, has not focused much on organizational culture, leaving the field open for others.
The new cognitive science of culture changes the game. In part II I will begin to explain why, and how, the new science can be put to use.
Some Further Reading:
1. Alvesson, M., & Sveningsson, S. (2008). Changing organizational culture: Cultural change work in progress. Abingdon: Routledge.
2. D’Andrade, R. (1995). The development of cognitive anthropology. Cambridge: Cambridge University Press
3. D’Andrade, R. (2008). A study of personal and cultural values: American, Japanese, and Vietnamese. New York: Palgrave Macmillan.
4. Descola, P. (2013). Beyond nature and culture. Chicago: University of Chicago Press.
5. Evans, (2004); Cognitive linguistics. Edinburgh University Press.
6. Evans, V. (2006). Lexical concepts, cognitive models and meaning-construction.Cognitive Linguistics17(4). DOI 10.1515/COG.2006.016
7. Gentner, D. & Colhoun, J. (2010). Analogical processes in human thinking and learning. In A. von Muller & E. Poppel (Series Eds.) & B. Glatzeder, V. Goel, & A. von Muller (Vol. Eds.), On thinking: Vol. 2. Towards a theory of thinking (pp. 35–48). Berlin: Springer-Verlag.
8. Handwerker, W. P. (2009). The origin of cultures. Walnut Creek, CA: Left Coast Press.
9. Hutchins, E. (2014). The cultural ecosystem of human cognition. Philosophical Psychology, 27(1), 34-49.
10. Hutchins, E. (2005). Material anchors for conceptual blends. Journal of Pragmatics,37, 1555-1577.
11. Hutchins, E. (1995). Cognition in the wild. Cambridge: MIT Press
12. Kronenfeld, D. (2018). Culture as a system: How we know the meaning and significance of what we do and say.New York: Routledge.
13. Langacker, R. (1987). Foundations of cognitive grammar, vol. 1, theoretical prerequisites. Stanford: Stanford University Press.
14. Martin, J., Frost, P. (2011). The organizational culture war games: a struggle for intellectual dominance. In M. Godwin and J. Hoffer Gittell (Eds.), Sociology of organizations: Structures and relationships(pp. 559-621). Thousand Oaks: Sage. Retrieved from: http://scholar.google.com/
15. Martin, J. (2002). Organizational culture: Mapping the terrain. Thousand Oaks: Sage. Ogbonna & Harris, (2002).
16. Ogbonna, E., & Harris, L.C., (2002). Managing organisational culture: insights from the hospitality industry. Human Resource Management Journal, 12(1), 33-53
17. Palmer, G. B. (1996). Toward a theory of cultural linguistics. Austin: University of Texas Press.
18. Quinn, N. (2005). How to reconstruct schemas people share, from what they say. In: N. Quinn (Ed.), Finding culture in talk.(pp: 35-84).New York: Palgrave. Smith (2003)
19. Shore, B. (1998). Culture in mind: Cognition, culture, and the problem of meaning. Oxford: Oxford University Press.
20. Smith, M. E. (2003). Changing an organisation’s culture: correlates of success and failure. Leadership & Organization Development Journal, 24(5), 249-261.
21. Stewart, J. (2010). Introduction. In: J. Stewart, O. Gapenne,& A. Di Paolo (Eds.). Enaction: Toward a new paradigm for cognitive science. Cambridge, MA: MIT Press.
22. Strauss, C. & Quinn, N. (1997). A cognitive theory of cultural meaning. Cambridge: Cambridge University Press.
23. Varela, F. J., Thompson, E., & Rosch, E. (1991). The embodied mind. Cambridge, MA: MIT Press.
25. Wood, J., & Vilkinas, T. (2004). Characteristics of chief executive officers views of their staff. Journal of Management Development,23(5), 469-478.
By “mainstream” I include most of the work emanating from business schools and large consultancies. It is not that some of this work is without merit. My issue is that little of it looks outside of the management literature or narrowly drawn client case studies to the fields referenced above. Most of the management scholarship on culture lives in an echo chamber, promulgating many of the theories shown to be problematic in this article.
 Known as the linguistic relativity problem outline by Edward Sapir and Benjamin Lee Whorf.
 This is not to deny language plays a role in constructing meaning. How groups make meaning is in part based on language, which in turn reflects some aspects of shared culture. But that is only part of the story. Simply positing a desired culture by identifying key words, phrases and slogans for people to adopt is nowhere near enough to actually change lived culture.
 Cultural models are the underpinnings of culture and are described in Part II.