New Book: “Five Constraints on Predicting Behavior” by Jerome Kagan!

Dear Commons Community,

Jerome Kagan, the Daniel and Amy Starch Emeritus Professor of Psychology at Harvard University, has a new book entitled,  Five Constraints on Predicting Behavior, that focuses on the relationship between brain activity and psychological processes.  As the title indicates, this book examines five conditions that serve to constrain inferences about the relation between the brain and behavior.  It is a brief book (170 pages minus the references) but delves deeply into the issues of the brain and the mind.  An understanding of the brain’s contributions to behavior, Kagan argues, requires investigators to acknowledge these five constraints in the design or interpretation of an experiment. Neuroscience is not my field and some of the explanations  about brain function are a bit murky for me, however, I found that Kagan makes most of the material approachable. Here is an excerpt from a review (see below for the full review) by Maura Pilotti that appeared in Metapsychology. 

“Five Constraints on Predicting Behavior is a seminal narrative that has the potential of reshaping the growing field of cognitive neuroscience. The principal reason is that his narrative addresses the obstacles that have prevented neuroscientists from understanding how time-bound, electrochemical states and changes, which define brain patterns across time, can translate into phenomena of perception, cognition, and behavior. Thus, it is not a narrative on any specific psychological phenomenon, but a call for action that offers a broad overview of the current state of the field. It does so by identifying critical hurdles in the practices through which understanding of the links between brain and mental states is pursued, as well as by offering encouraging solutions that require a paradigmatic shift in the way evidence is collected and interpreted.

In the introduction to his book, Kagan mentions as culprits theories and methods that are currently inadequate to uncover the formula(s) by which spatially defined electrochemical changes become perceptions, thoughts, feelings, and actions. He also mentions the scarce funding of young investigators whose efforts can move the scientific enterprise in the right direction. He then directs his attention to five critical issues, which constitute the fabric of his narrative. From his standpoint, these issues represent serious obstacles in the quest to understand how brain matters give rise to psychological and behavioral phenomena. Kagan notices that, in order to generate reliable and valid interpretations, (1) the context of data collection, (2) the role of expectations, and (3) the quality of the sources of evidence cannot be disregarded. He also mentions that (4) selective consideration of co-variation, rather than patterns of change, is counterproductive, and warns of the (5) dangers of blindly borrowing the language of the social and behavioral sciences to discuss brain-related phenomena.

All things considered, Karan’s narrative is informative, refreshing and enlightening.”

I agree with Dr. Pilotti’s review completely and would add that there are kernels in Kagan’s book that are important for his main audience but also for social science researchers.   For example,  on page five as he sets up the rationale for the book, Kagan quotes Niels Bohr that “the validity of every conclusion depends on its source of evidence.”  Bohr was referring to the quantum controversy as to whether light was a wave or particle but Kagan application to his subject is most appropriate.

In sum, a good read!

Tony

——————————————————–

Metapsychology  

Review by Maura Pilotti, PhD
Feb 20th 2018 (Volume 22, Issue 8)

Five Constraints on Predicting Behavior, written by Jerome Kagan, is a seminal narrative that has the potential of reshaping the growing field of cognitive neuroscience. The principal reason is that his narrative addresses the obstacles that have prevented neuroscientists from understanding how time-bound, electrochemical states and changes, which define brain patterns across time, can translate into phenomena of perception, cognition, and behavior. Thus, it is not a narrative on any specific psychological phenomenon, but a call for action that offers a broad overview of the current state of the field. It does so by identifying critical hurdles in the practices through which understanding of the links between brain and mental states is pursued, as well as by offering encouraging solutions that require a paradigmatic shift in the way evidence is collected and interpreted.

In the introduction to his book, Kagan mentions as culprits theories and methods that are currently inadequate to uncover the formula(s) by which spatially defined electrochemical changes become perceptions, thoughts, feelings, and actions. He also mentions the scarce funding of young investigators whose efforts can move the scientific enterprise in the right direction. He then directs his attention to five critical issues, which constitute the fabric of his narrative. From his standpoint, these issues represent serious obstacles in the quest to understand how brain matters give rise to psychological and behavioral phenomena. Kagan notices that, in order to generate reliable and valid interpretations, (1) the context of data collection, (2) the role of expectations, and (3) the quality of the sources of evidence cannot be disregarded. He also mentions that (4) selective consideration of co-variation, rather than patterns of change, is counterproductive, and warns of the (5) dangers of blindly borrowing the language of the social and behavioral sciences to discuss brain-related phenomena.

All things considered, Karan’s narrative is informative, refreshing and enlightening. He begins his probing analysis with the acknowledgment that data collection occurs in context and that the latter shapes the phenomena under observation and thus their interpretation(s). The term context is used to refer to the physical and social properties of the situation where the data of a study are collected, including the characteristics of the participants, the range of stimuli considered, and the features of the data collection procedures adopted by the investigator. Kagan argues that “context effects” limit the generalizability of the data and related interpretations. As such, they cannot be easily dismissed. On the contrary, context effects are intended to be carefully investigated not only to determine the extent to which phenomena and their interpretations may generalize to a variety of settings and circumstances besides those that produce them, but also to identify whether specific factors in such settings and circumstances shape the observed phenomena to the point at which very different configurations develop. Namely, he advocates the practice of systematic replication as a means by which explanations can be validated and their applicable range established. In doing so, he illustrates several helpful as well as compelling examples in which the context of observation matters. Kagan’s argument is particularly persuasive since scientists involved in the Reproducibility Project, led by Brian Nosek, have succeeded in replicating the findings of only a small number of famous psychological studies published in top journals (Francis, 2012; Open Science Collaboration, 2015).

Kagan’s comments on the relevance of the context of observation fit well with another concern he raises regarding the role played by expectations in shaping brain profiles. He argues that a participant’s response to a stimulus in either a lab or the field is the complex outcome of the information gathered from the stimulus and the context of occurrence, as well as prior knowledge that leads the participant to anticipate that stimulus or be surprised by its occurrence. Expectations may merely affect the intensity of a response, or shape its meaning and nature entirely. Thus, Kagan advises scientists whose goal is to explain how brain changes lead to perception, cognition, and action, to consider the properties of the situation in which data collection occurs in relation to the selected participants. Failure to do so may not only limit the generalizability of particular findings, but also question their interpretation and cast doubt on their validity and reliability. The same liabilities may emerge if the sources of the data collected, which determine the validity of the evidence gathered, are not given sufficient consideration, and are thus discounted or ignored. Kagan reminds the reader that the validity of interpretations, including the range of contexts, events, and individuals to which they apply, relies, first and foremost, on the validity of the data collected. As examples, he mentions scientists’ over-reliance on data gathered through subjective report measures, such as questionnaires, either to classify human psychological diversity into personality traits or to explain the impact of early experiences on later psychological states. People’s recollections of past events, including habitual behaviors and cognitions, are neither the events themselves nor the untarnished duplicates of such events. Memories, for instance, are shaped by the context of retrieval, including current perceptions (e.g., retrieval cues), emotional states, and intentions, and may even be illusory (Berntson & Cacioppo, 2009; Tulving & Craik, 2000).

The remaining obstacles to the progress of cognitive neuroscience are equally relevant. Kagan overviews several prototypical instances of scientists’ predilection for uncovering one-to-one relationships involving a variety of psychological, behavioral, and brain phenomena. Simplistic assumptions may span from mere co-variation to the postulation of a cause-effect relationship. The simplicity of a one-to-one relationship, whereby a psychological phenomenon is attributed to a single identifiable source, may be deceptively comforting. In fact, the principle of parsimony appears to be satisfied in a world forcibly made simple, and the chimera of straightforward applications and/or solutions can blind even the keenest minds. Kagan not only advocates attention to patterns of variables and non-linear relationships, but artfully demonstrates the value of data collection focused on patterns of measures representing likely causes and related outcomes. Lastly, Kagan acknowledges that the narratives of cognitive neuroscientists tend to be distorted by their borrowing terms and expressions from a language that may fit psychological and behavioral phenomena, but is inadequate to represent brain assemblies and patterns. He argues that if scientists want to overcome their current inability to explain how mental states emerge from brain patterns, then the latter need to be described through a language that does justice to their unique properties. To demonstrate this point, he explores a variety of cases where the tendency to attribute psychological properties to brain events can generate misleading conclusions.

Taken as a whole, Kagan’s narrative is both engaging and carefully crafted for a broad spectrum of readers who are interested in the progress of the complex field of cognitive neuroscience. Not only can Kagan’s advice and guidance promote the advancement of the field, but his narrative can serve as a call for action. When expertise and imagination meet intellectual humility, the path to progress can be envisioned. Kagan has artfully plowed it.

References

Berntson, G. G., & Cacioppo, J. T. (Eds.). (2009). Handbook of neuroscience for the behavioral sciences.(Vol. 2). Hoboken, NJ: John Wiley & Sons.

Francis, G. (2012). Publication bias and the failure of replication in experimental psychology. Psychonomic Bulletin & Review19(6), 975-991.

Open Science Collaboration. (2015). Estimating the reproducibility of psychological science. Science349(6251), aac4716.

Tulving, E., & Craik, F. I. (Eds.). (2000). The Oxford handbook of memory. Oxford: Oxford University Press.

Comments are closed.