What do you think?
Rate this book
The Ambiguities of Experience
The first component of intelligence involves effective adaptation to an environment. In order to adapt effectively, organizations require resources, capabilities at using them, knowledge about the worlds in which they exist, good fortune, and good decisions. They typically face competition for resources and uncertainties about the future. Many, but possibly not all, of the factors determining their fates are outside their control. Populations of organizations and individual organizations survive, in part, presumably because they possess adaptive intelligence; but survival is by no means assured. The second component of intelligence involves the elegance of interpretations of the experiences of life. Such interpretations encompass both theories of history and philosophies of meaning, but they go beyond such things to comprehend the grubby details of daily existence. Interpretations decorate human existence. They make a claim to significance that is independent of their contribution to effective action. Such intelligence glories in the contemplation, comprehension, and appreciation of life, not just the control of it.―from The Ambiguities of Experience In The Ambiguities of Experience , James G. March asks a deceptively simple What is, or should be, the role of experience in creating intelligence, particularly in organizations? Folk wisdom both trumpets the significance of experience and warns of its inadequacies. On one hand, experience is described as the best teacher. On the other hand, experience is described as the teacher of fools, of those unable or unwilling to learn from accumulated knowledge or the teaching of experts. The disagreement between those folk aphorisms reflects profound questions about the human pursuit of intelligence through learning from experience that have long confronted philosophers and social scientists. This book considers the unexpected problems organizations (and the individuals in them) face when they rely on experience to adapt, improve, and survive. While acknowledging the power of learning from experience and the extensive use of experience as a basis for adaptation and for constructing stories and models of history, this book examines the problems with such learning. March argues that although individuals and organizations are eager to derive intelligence from experience, the inferences stemming from that eagerness are often misguided. The problems lie partly in errors in how people think, but even more so in properties of experience that confound learning from it. "Experience," March concludes, "may possibly be the best teacher, but it is not a particularly good teacher."
168 pages, Hardcover
First published April 1, 2010
29 people are currently reading
149 people want to read
About the author
James G. March
44 books30 followersJames Gardner March was an American sociologist who was professor at Stanford University and the Stanford Graduate School of Education, best known for his research on organizations, his A behavioral theory of the firm and organizational decision making.
Ratings & Reviews
Friends & Following
Create a free account to discover what your friends think of this book!
Community Reviews
Displaying 1 - 8 of 8 reviews
December 9, 2019
Learning is conspicuous in human existence.
By learning, we mean that people change expectations/understandings based on the observed outcome of actions, using two mechanisms: low-intellect (simple replication of success) and high-intellect (fitting a causal model to experience).
These learning mechanisms lead to improvement under certain conditions
- Relatively constrained activities + repeated frequently, ex: production learning curves, continuing relationships among people, and the development of technical/artistic skills
- This means a high signal to noise ratio and a large sample size
- Performance is improved but not necessarily optimal and has limited generalisability
- Same as for machines!
Novelty generation is required for the long-run effectiveness of learning from experience. It is threatened by effective learning, but can also be engineered.
In other conditions, experience is not a good teacher (by that we mean reliably improving performance).
- Vividness of direct experience leads to exaggeration of the information content therein
- The inherent ambiguity of experience makes it hard to learn from.
1. Complex causal structure makes unjustified conclusions, superstitious associations, misleading correlations, tautological generalisations, and systematic biases all more likely!
2. History is but a single instance from a distribution, which is a poor representation of the distribution. Counterfacturals can only be generated using imagination...
3. History is path-dependent: one earlier choice could influence downstream outcomes
4. History is not "pure observations", it is stories that we tell ourselves, concocted for a purpose
5. History has a terribly small sample size
6. Given such a complex system, effect detection requires holding other things constant and make a big enough change, but we tend to change many things simultaneously and to make small changes...
The ambiguity of experience means that the same experience can be interpreted in many different ways. Also, there is a tendency to use a catch-all term to explain everything that is not explained, providing an appearance of explanatory power. Ex: personality in psychology, culture in sociology/anthropology, uplift in power price forecasting.
The actual state of knowledge in oranisations:
1. A collection of theorems derived from a few elementary propositions about human behaviour
2. An engineering conception of knowledge as opposed to a scientific one: Scientific: understand system well enough to anticipate outcome of any possible antecedents vs. Engineering: seeks set of antecedents to produce a particular outcome
3. Instead of seeking to understand a complex word, orgs try to manage it using tools such as contracts, insurance, and hedges
But, in those cases, interpretation of experience has purposes other than increasing our predictive power, notably the creation and decoration of irrelevant understanding (our intellectual plumage) to show off the grace of our storytelling and the elegance of our models, an essential/fundamental human activity.
By learning, we mean that people change expectations/understandings based on the observed outcome of actions, using two mechanisms: low-intellect (simple replication of success) and high-intellect (fitting a causal model to experience).
These learning mechanisms lead to improvement under certain conditions
- Relatively constrained activities + repeated frequently, ex: production learning curves, continuing relationships among people, and the development of technical/artistic skills
- This means a high signal to noise ratio and a large sample size
- Performance is improved but not necessarily optimal and has limited generalisability
- Same as for machines!
Novelty generation is required for the long-run effectiveness of learning from experience. It is threatened by effective learning, but can also be engineered.
In other conditions, experience is not a good teacher (by that we mean reliably improving performance).
- Vividness of direct experience leads to exaggeration of the information content therein
- The inherent ambiguity of experience makes it hard to learn from.
1. Complex causal structure makes unjustified conclusions, superstitious associations, misleading correlations, tautological generalisations, and systematic biases all more likely!
2. History is but a single instance from a distribution, which is a poor representation of the distribution. Counterfacturals can only be generated using imagination...
3. History is path-dependent: one earlier choice could influence downstream outcomes
4. History is not "pure observations", it is stories that we tell ourselves, concocted for a purpose
5. History has a terribly small sample size
6. Given such a complex system, effect detection requires holding other things constant and make a big enough change, but we tend to change many things simultaneously and to make small changes...
The ambiguity of experience means that the same experience can be interpreted in many different ways. Also, there is a tendency to use a catch-all term to explain everything that is not explained, providing an appearance of explanatory power. Ex: personality in psychology, culture in sociology/anthropology, uplift in power price forecasting.
The actual state of knowledge in oranisations:
1. A collection of theorems derived from a few elementary propositions about human behaviour
2. An engineering conception of knowledge as opposed to a scientific one: Scientific: understand system well enough to anticipate outcome of any possible antecedents vs. Engineering: seeks set of antecedents to produce a particular outcome
3. Instead of seeking to understand a complex word, orgs try to manage it using tools such as contracts, insurance, and hedges
But, in those cases, interpretation of experience has purposes other than increasing our predictive power, notably the creation and decoration of irrelevant understanding (our intellectual plumage) to show off the grace of our storytelling and the elegance of our models, an essential/fundamental human activity.
February 4, 2018
A useful corrective to what March might refer to as "managerial overconfidence." March is widely read in the relevant philosophical and managerial literature, and points to many interesting books and studies on the subject
Throughout the book, the author presents a hardcore skeptics view that focuses on the many ways organizations can infer the wrong lessons and draw incorrect conclusions from experience. Sometimes the argument seems to border on epistemological nihilism, in that knowledge is so difficult to ascertain as to be impossible, but then in the next paragraph he'll make a tactical retreat from such an extreme position.
March sums up his argument toward the end of the book as such:
"If there is one lesson to be gleaned from the explorations in this book, it is that learning from experience is an imperfect instrument for finding the truth. Much of organizational and managerial life will produce vividly compelling experiences from which individuals and organizations will learn with considerable confidence, but the lessons that they learn are likely to be incomplete, superstitious, self-confirming, or mythic.
They will characteristically lead to suboptimal choices and are unlikely to yield valid characterizations of the causal processes underlying the experiences. Experience is likely to generate confidence more reliably than it generates competence and to stop experimentation too soon."
Throughout the book, the author presents a hardcore skeptics view that focuses on the many ways organizations can infer the wrong lessons and draw incorrect conclusions from experience. Sometimes the argument seems to border on epistemological nihilism, in that knowledge is so difficult to ascertain as to be impossible, but then in the next paragraph he'll make a tactical retreat from such an extreme position.
March sums up his argument toward the end of the book as such:
"If there is one lesson to be gleaned from the explorations in this book, it is that learning from experience is an imperfect instrument for finding the truth. Much of organizational and managerial life will produce vividly compelling experiences from which individuals and organizations will learn with considerable confidence, but the lessons that they learn are likely to be incomplete, superstitious, self-confirming, or mythic.
They will characteristically lead to suboptimal choices and are unlikely to yield valid characterizations of the causal processes underlying the experiences. Experience is likely to generate confidence more reliably than it generates competence and to stop experimentation too soon."
May 16, 2021
Excellent concise treatment of the many problems inherent to experiential learning by individuals and organizations. As March asserts, "Experience may be the best teacher, but it is not a particularly good teacher." Discovered this little book through a footnote in Soyer and Hogarth's far more popularized recent treatment of the topic, "The Myth of Experience." I recommend both for understanding how our assumptions about the value of experiential learning, while appropriate when applied to tasks defined by relatively limited complexity and frequent occurrence, mostly falls apart when involving highly dynamic and casually multivariate real world problems and situations (like warfare). There is plenty of interesting thought in here concerning the influence of "mythic frames" and storytelling on human rationalizations of lived and vicarious experience, as well as an underscoring of the "flexibility of interpretation" that will provide much food for thought for historians.
January 21, 2020
Found this book when I feel my job is meaningless and wanted to know what work experience meant, and what values it could bring. A couple years has passed and I finally got some time to flip through this book. I don’t think I found the answer I was looking for, but I do feel I need to come back to read it again some years later with more peaceful mind. Maybe I will have new understandings then.
December 4, 2016
From the material of Cornell University lectures. Written a bit academically (not on purpose, the author did his best. bearable.)
The point he wants to prove here is that learning from experience is better suited for narrow and predictable fields. Domain-specific expertise. And may be inapplicable to situations with high level of noise over signal, as in business & politics.
'Knowledge is sought about experience, but experience is ambiguous and its causal structure complex. Identities are ambiguous, and the ways in which they apply to particular situations are intricately multifaceted. The development and elaboration of stories and models of experience are a major part of the documentation of management. Participants are asked what happened and why it happened, and research often consists primarily in tabulating the stories told by storytellers.'
In the middle of the book, I`m not sure: is he hard or social scientist?
'Thus, life in organizations is marked by a quest for stories and models of maximum comprehensible complexity, an interpretation of life that is simultaneously suitably interesting and suitably comprehensible to signal intelligence, and for the development of individuals who can expound such interpretations.'
Section on novelty looks like distilled evolutionary biology, e.g. Koonin. However, reduced and less powerful. That's the source, I think.
'Learning from experience requires experimentation for its long-run effectiveness but tends to extinquish it. Novelty is habitually vulnerable to effective learning.'
'Experience may possibly be the best teacher, but it is not particularly good teacher'
Looking forward to this topic being explored by a younger (less experienced) writer. It`s more of a business than psychology book. Worth recommending.
The point he wants to prove here is that learning from experience is better suited for narrow and predictable fields. Domain-specific expertise. And may be inapplicable to situations with high level of noise over signal, as in business & politics.
'Knowledge is sought about experience, but experience is ambiguous and its causal structure complex. Identities are ambiguous, and the ways in which they apply to particular situations are intricately multifaceted. The development and elaboration of stories and models of experience are a major part of the documentation of management. Participants are asked what happened and why it happened, and research often consists primarily in tabulating the stories told by storytellers.'
In the middle of the book, I`m not sure: is he hard or social scientist?
'Thus, life in organizations is marked by a quest for stories and models of maximum comprehensible complexity, an interpretation of life that is simultaneously suitably interesting and suitably comprehensible to signal intelligence, and for the development of individuals who can expound such interpretations.'
Section on novelty looks like distilled evolutionary biology, e.g. Koonin. However, reduced and less powerful. That's the source, I think.
'Learning from experience requires experimentation for its long-run effectiveness but tends to extinquish it. Novelty is habitually vulnerable to effective learning.'
'Experience may possibly be the best teacher, but it is not particularly good teacher'
Looking forward to this topic being explored by a younger (less experienced) writer. It`s more of a business than psychology book. Worth recommending.
January 20, 2016
Experiential learning is supposed to be the most powerful form of learning, but as March explains, "Experience may possibly be the best teacher, but it is not a particularly good teacher."
July 4, 2016
Interesting. But not at all what I was expecting.
February 7, 2021
Discussed the limitation, pros, and cons of experience at a very high level perspective.
Displaying 1 - 8 of 8 reviews