Where Are The Experts? Why Aren’t They Here?
Almost ten years ago, I returned to full-time education for two years, completing a post-grad in psychology. In celebration of this enriching experience, I am sharing some of my favourite essays on themes of psychology relevant to today’s culture. The essay below (original title: How do experts use their knowledge differently from novices?) is a favourite because it was course work for a professor at TCD who was truly outstanding – check out the work of Ruth Byrne here – and because its meaning resonated with my experience in professional life: Why is it important to have the experienced people in the key strategic thinking sessions? Why should the facilitator make sure that those with the most experience (not necessarily the most seniority) be fully heard? Why do junior people in the room offer freshness of perspective while they parlay their intelligence into expertise?
Knowledge is power (Bacon, 1597). This reductive aphorism is perhaps only partially true in the light of this essay, which explores the proposition that it is the optimal usageof knowledge which is most powerful. The essay seeks to trace the usage of knowledge by experts as compared to novices. In so doing, it outlines the meaning of knowledge, expertise, and novice-hood – and then moves to treat the two major theories (Chunking Theory and Template Theory) which help to explain how experts and novices act so differently with knowledge. Significant time is spent in the world of chess, considering experimentation by thought leaders such as De Groot (1965) and Gobet (1998). Other experimentation, most notably in the medical field, has added to our understanding of the expert and how knowledge is processed differently. Finally, the author stands back and considers what the theory and experimentation can and cannot tell us about how experts and novices use their knowledge.
PROLOGUE: DISTINGUISHING NOVICES FROM EXPERTS – A FOCUS ON PROCESS, NOT OUTCOMES
i) The novice heroine
On the 26th December 2004, ten-year-old Tilly Smith was on a seaside holiday in Thailand, when she observed a rapid withdrawal of seawater from the beach, accompanied by a gurgling sound and extensive frothing. Tilly connected what she was witnessing with her recent school lesson regarding the prediction of natural disasters. She concluded that a tsunami was about to occur. She immediately reacted, warned her family and the authorities, who then cleared the beach. Her rapid action on the day of the Asian Tsunami is attributed to saving the lives of up to 100 people (United Nations International Strategy for Disaster Reduction Secretariat, 2005).
ii) The expert hero
At 3:37:01, on the 15thJanuary 2009, pilot Chesley (Sully) Sullenberger’s aircraft was struck by birds in both engines within seconds of its departure from La Guardia, New York, causing complete engine failure. He promptly took control of the aircraft and considered the alternatives given the aircraft’s low altitude. At 3:28:05, Captain Sullenberger decided that ditching in the Hudson River was his best alternative(Sullenberger, 2009), thereby discounting a return to La Guardia or an onward glide to a local New Jersey airstrip. Famously, he did just that, and all 155 people on board were rescued. Remarkably, this critical decision was made in 64 seconds, and under the most extreme of pressures. Cockpit recordings capture an atmosphere of professional calm throughout that period (Miracle on the Hudson, 2009).
These two decisions, based on each actor’s available knowledge, have been justly lauded as heroism in the media. From a cognitive perspective, however, they display quite different characteristics. In the case of Tilly Smith, her actions can be ascribed to the ‘beginner’s luck’ of a novice who makes a linear connection (i.e. drawback of water and gurgling are the ‘rules’ to expect a tsunami) resulting in a good decision. Though her actions were likely an act of creative intelligence, they were not the result of expertise. Captain Sullenberger, in contrast, had 19,663 hours flying experience on 15thJanuary 2009 (Sullenberger, 2009). The evidence suggests a heavy reliance on implicit knowledge, strategic clarity and heuristic thinking as he rapidly comprehended a catastrophe and took action. The important distinction is in the processes by which knowledge was used in these instances, and is at the heart of this essay.
THE NATURE OF KNOWLEDGE
Knowledge is a fund of skill and understanding in a given domain, built through learning and experience (Eysenck & Keane, 2010). Knowledge can be seen as a currency of the human experience – the culmination of perception and categorisation which enables us to operate in the world (Bruner, 1979). The acquisition of knowledge can be both implicit and explicit. The former refers to learning gained without conscious recollection (Eysenck & Keane, 2010) and can be compared to the process of osmosis. Explicit knowledge acquisition differs, in that the learning mechanism involves conscious encoding of information. Importantly, knowledge can also be accessed and used both implicitly and explicitly (Eysenck & Keane, 2010).
THE NATURE OF EXPERTS AND EXPERTISE
Expertise is highly skilled, competent performance in a task or domain (Eysenck & Keane, 2010). It is characterised by: a) ability at a high level; b) the generation of a result; c) specialisation (one cannot be an expert in everything). The concept of expertise is relative: there are experts and novices in every field, in a continuum of proficiency.
A novice is characterised as someone unskilled in a domain. This should not be confused with a lack of either intelligence or potential, as an expert and novice can often share equal IQ (Eysenck & Keane). Rather, the salient feature of the novice is a lack of exposure in a given domain. Cognitive psychology asserts that novices and experts think, process and decide differently in several important ways (Gladwell, 2007).
THE CONTRIBUTION OF CHESS TO THEORY AND UNDERSTANDING OF EXPERT VERSUS NOVICE PROCESSES
The study of chess players is an ideal landscape to examine expertise at work. It is considered an unparallel laboratory, as the learning process and the degree of ability obtained can be measured exactly (Russkin-Gutman, 2009). Further, chess exhausts the limits of explicit, rational thinking and demands strategic choice, as the number of possible chess games is 10120– a figure greater than the number of atoms in the universe (Kasporov, 2010).
ii) De Groot – pioneering experimenter
Adriaan de Groot’s influential book ‘Thought and choice in chess’ (De Groot, 1965) sets out the foundation for much of the theory of expertise. A Dutch Grand Master himself, he sought to account for the differences in the processes of chess experts and novices (or ‘patzers’) through experimentation and theory (De Groot, 1978, p1).
Using ‘talk aloud’ protocols, De Groot (1965) showed 10 chess players (5 Grand Masters; 5 Candidate Masters) the same board position and asked to determine the best move for White. Surprisingly, both groups appeared to act in the same manner, using a similar number of moves, depth-of-search and positions visited, though the Grand Masters played better moves than lesser experts. This led De Groot to suspect that the Grand Masters’ superiority lay not in the macrostructure of search processes, but rather in their perception and memory of positions (De Groot, 1965). Although De Groot did not carry out his research on novices, the difference found between these two groupings is helpful in a directional sense (see Bilalic & McLeod, 2008 for a discussion of how de Groot’s work is regularly misrepresented in the literature).
De Groot’s experimentation (De Groot, 1965) also shed light on how memory operates with Grand Masters. In presenting Grand Masters and Chess Experts with a loaded, non-random board (20-24 pieces) for up to 15 seconds, he discovered that Grand Masters had 91% correct recall of positions, compared to 41% for Chess Experts, implying an exponential improvement for Grand Masters. De Groot, after eliminating the possibility of better short-term memory, asserted that the delta reflected differences in stored chess information – the ability to chunk information efficiently, due to the deeper meaning encoded in each chunk (de Groot, 1965).
iii) Development of Chunking Theory
Chase & Simon (1973) developed Chunking Theory based on De Groot’s thinking, and determined that expertise depends both on the availability in memory of information about a larger number of recurring patterns and on the availability of strategies for selective search. In an experiment measuring the amount of time better and weaker players took to examine and recreate a chess board, expert players were found to chunk greater amounts (2.5 pieces) versus weaker players (1.9 pieces), and did so in less time, by use of what the authors termed the mind’s eye(Chase & Simon, 1973). Chunking theory can be seen as an extension of Miller’s magic number 7±2, the size of typical working memory, with experts operating significantly beyond novices (Miller, 1956), and with the ability of placing upwards of 1000,000 chunks into long term memory (Eysenck & Keane, 2010).
iv) Chunking theory critiqued, and the emergence of Template Theory
Swiss chess master and psychologist, Fernand Gobet is at the heart of current thinking on Expertise Theory. Gobet and Waters (2003) criticised Chunking Theory, stating that it failed to relate chunk level mechanisms with the representations used by experts; further, chunking theory predicted that it would take a long time to encode chess positions – a prediction that turns out to be false (Eysenck & Keane, 2010).
Template Theory was proposed as a more accurate representation of how experts use their knowledge (Gobet, 1998). Rather than rely on the photographic-style memory of chunking, templates consist of both core (unchanging information chunks) and slots (variable information). Template theory combines the concept of chunking with a retrieval structure to create a more detailed model for expertise. The strength of the theory is its ability to connect low-level and high-level information (Eysenck & Keane, 2010).
v) Template Theory being tested
Template Theory has proven rich territory as it makes several clear and testable predictions with strong empirical results. Gobet and Clarkson’s (2004) research in exemplars is a case in point. In an experiment of 12 participants at three chess levels, the number and size of chunks used in a copy and recall taskfor both game and random boards was measured. It was found that no more than three chunks were replaced in the recall tasks (as predicted by Template Theory) and chess masters replaced very large chunks – up to 15 pieces (also in line with Template Theory). These findings signalled the end of Chunking Theory – and a breakthrough in modelling how experts use their knowledge so effectively. However, Template Theory is not without imperfections. It does not seem to account adequately for how adaptive, as opposed to routine, expertise is employed by experts (Eysenck & Keane, 2010). Further, it does not adequately account for the speed with which Grand Masters operate, nor does it fully account for the unique cognitive nexus that chess inhabits – a place, to paraphrase Kasporov (2007), where science and art fuse in the mind, and are refined and improved by experience.
WHAT MAKES EXPERTS DIFFERENT?
i) Practice: the hallmark of experts
Expertise is built through time perfecting understanding of a chosen field. A major thrust of Gladwell’s (2008) ‘Outliers’ is the thesis that experts are not born but made. Gladwell claims that experts are often subject to Ross’ (1975) Fundamental Attribution Error, whereby their success is ascribed to personal predispositions (unjustifiably) rather than the environment. In a meta-analyses of a wide range of expert domains (typing, juggling, music, chess, medicine), Ericsson and Lehmann (1996) found that the highest levels of human performance can only be attained after ten years of extended deliberate practice, including feedback and correction in the pursuit of perfection. Through such training experts are believed to circumvent the general limits of reaction time, and increase distinctive memory skills required for the service of their craft.
It is interesting to note that experts themselves often accentuate circumstance over intellectual giftedness in reasoning why they possess expertise:
‘If I have seen further it is by standing on the shoulders of giants’(Isaac Newton,1675)
‘Genius is one percent inspiration, ninety-nine percent perspiration’(Thomas Edison,1932)
What sets experts apart is how they use their accumulated knowledge, rather than their innate ability to garner knowledge in the first place (Gobet, 1998).
ii) Depth: categorisation is on a deeper level
Chi, Feltovich & Glaser (1981) compared how 8 advanced PhD students (experts) and 8 undergraduates (novices) categorised physics problems, and found a distinct difference in sorting techniques. Experts tended to take more time in appraising the situation, then abstracted the issues, creating sets of ostensibly different problems with underlying similarities. Novices were speedy to act, but were linear in their approach, categorising based on literal features. Chi et al. concluded that experts can access implicit, procedural knowledge and schemas to make their assessments at a deeper level, whilst novices operate from a declarative and explicit knowledge base, and hence displayed more shallow processing.The finding is similar to Gladwell’s (2008) assertion: novices tend to think tactically whilst experts tend to think strategically.
THE CONTRIBUTION OF MEDICAL EXPERIMENATATION TO THE UNDERSTANDING OF EXPERT VERSUS NOVICE PROCESSES
Medical expertise, especially in the visual specialities, has shed further light on the dynamics of expertise in action. Two findings are noteworthy:
i) Experts hone in on the strategic areas of interest faster, and this speed is correlated to positive outcome
Kundel, Nodine, Conant and Weinstein (2007) monitored the patterns of eye movement amongst medics of differing levels of expertise (3 full-time mamographers; 1 attending radiologist; 2 mamographer fellows; 3 radiologist residents) in examining mammograms. The more expert participants fixated on the cancer zones faster (under 1 second versus 1.3 seconds mean) and this speed was positively correlated with correct performance. The authors conclude that experts’ holistic global processes are more efficient than the search-and-find techniques of relative novices. These results build on Chi’s (1981) findings, above.
ii) Medical Experts use speedy exemplar-based strategies, not rules
Kulatunga-Moruzi, Brooks and Norman (2004) measured expert and non-expert medics’ processing strategies in dealing with skin lesions. Three levels of experts were assessed (8 dermatologists; 12 general practitioners and 10 resident doctors). Their findings are somewhat counter-intuitive: medical experts’ correct diagnoses reduce (although still superior to novices at all stages) when confronted with verbal information BEFORE seeing visual information (78%) versus seeing visual information first (90% correct). The authors assert that because experts use the exemplar method, verbal descriptions tend to interfere with this strategy. This finding supports Template Theory. In short, information must not just be ‘good’, but relevant to the user – and expert needs are different to novice needs, as the former use exemplars and the latter operate in a more rule-bound and explicit environment.
The literature compellingly demonstrates differences in the usage of knowledge by experts versus novices: their access to implicit as well as explicit knowledge, their ability to observe deep structures over shallow resemblances, their ability to chunk and retain information at a higher level of operation, their ability to remain flexible and ‘dance in the moment’ when confronted with new circumstances.
It is also striking what is NOT asserted: experts are not substantially more intelligent than novices, nor are many of their behaviours, such as depth of search, necessarily different (De Groot, 1965). The key seems to be in the way their experience has been encoded, moving from explicit to implicit knowledge, and accessed heuristically without delay. Because experts have often ‘seen it all before’, they know intuitively how to act.
This exploration of expert versus novice processes raises some interesting issues:
In practical environments such as healthcare (Kulatunga-Moruzi et al., 2004) the provision of optimal, not maximal, information can be a life or death distinction. The sharpest understanding of how experts act will have effects in the real world.
On a more theoretical level, Template Theory still has some gaps to fill, most especially in accounting for adaptive expertise which does not directly use core chunks, as the theory predicts. The present author, in reviewing the literature, holds a suspicion that Template Theory has been retro-fitted to suit the evidence of the chessboard, rather than one which sits comfortably within our understanding of memory and intelligence.
The world seems to benefit exponentially from experts – their critical mass of experience seems to ‘pay-back’ at impressively high levels. De Groot’s (1965) early experimentation hints at this, and intuitive observation confirms it. As society moves towards veneration of the generalist (e.g. emphasis on job rotation / promulgation of broad, liberal education) perhaps something will also be lost?
The present author wishes to close by according respect and deference to novices. They too have their role. Not just the once-off heroism of young Tilly Smith, but also the ability to stand up and demand of experts to forget their experience and consider a different way. After all, cognitive psychology has been the beneficiary of such actions. Noam Chomsky (1967), a relative novice, stood up and criticised B. F. Skinner, the sublime behavioural expert, with regard to the behaviourist approach to language learning. The result was a revolution in language theory and the beginning of a new era. The world needs both experts and novices.
Bacon, F. (Sir) (1597 / 2007) Religious Meditations: Of Heresies. In Francis Bacon: The Major Works. Oxford: Oxford University Publishers
Bilalic, M. & McLeod, P. (2008). Expert and ‘‘novice’’ problem solving strategies in chess: Sixty years of citing de Groot. Thinking and Reasoning, 2008, 14 (4), 395 – 408
Bruner, J. (1979).On Knowing: Essays for the Left Hand. Boston: Harvard University Press
Chase, W.G., & Simon, H.A. (1973). The mind’s eye in chess. In W.G. Chase (Ed.), Visual information processing. New York: Academic Press
Chi, M.T.H., Feltovich, P.J., Glaser, R. (1981) Categorization and Representation of Physics Problems by Experts and Novices. Cognitive Science. 5, 121-152
Chomsky, N. (1967) A Review of B. F. Skinner’s Verbal Behavior, In Leon A. Jakobovits and Murray S. Miron (eds.), Readings in the Psychology of Language. New York: Prentice-Hall, 1967, pp. 142-143
De Groot, A. (1965) Thought and choice in chess; 2ndedition. Amsterdam:Mouton De Gruyter
Edison, T. (1932) Spoken statement (c. 1903); published in Harper’s Monthly(September 1932)
Ericsson, K. & Lehmann, A. (1996). Expert and exceptional performance: Evidence of Maximal Adaptation to Task Constraints Annual Review of Psychology, 47 (2), 273-305
Eysenck, M. & Keane, M. (2010). Cognitive Psychology – a student’s handbook, 6th edition.New York: Psychology Press
Gladwell, M. (2008) Outliers: the story of success. New York: Little, Brown & Company
Gobet, F. (1998). Expert memory: a comparison of four theories. Cognition. 66, 115-152
Gobet, F. (1997) A Pattern-recognition Theory of Search in Expert Problem Solving. Thinking and reasoning3 (4), 293-313
Gobet, F., & Clarkson, G. (2004). Chunks in expert memory: Evidence for the magical number four…or is it two? Memory.12, 732-747
Kasparov. G (2010). The Chess Master and the Computer. New York Review of Books,57(2), 7-12
Kasporov, G. (2007) How Life Imitates Chess: Making the Right Moves, from the Board to the Boardroom. New York: Bloomsbury
Kulatunga-Moruzi, C., Brooks, L. & Norman, G. (2004) Using Comprehensive Feature Lists to Bias Medical Diagnosis. Journal of Experimental Psychology: Learning, Memory, and Cognition. 30 (3), 563–572
o Kundel, H., Nodine, C, Conant, E, & Weinstein, S. (2007) Holistic Component of Image Perception in Mammogram Interpretation: Gaze-tracking Study. Radiology. 242, 396-402
Klein, G. (1998). Sources of power: How people make decisions. Cambridge, MA: MIT Press
Lockhart, R. & Craik, F. (1990).Levels of Processing: a retrospective commentary on a framework for Memory Research. Canadian Journal of Psychology, 44(1), 87-112
Miller, G. A. (1956). The magical number seven, plus or minus two: Some limits on our capacity for processing information. Psychological Review, 63 (2): 81-97
Newton, I. (1675) Letter to Robert Hooke, February 5, 1675
Ross, L. (1977). The intuitive psychologist and his shortcomings: Distortions in the attribution process. ‘In L. Berkowitz (Ed.), Advances in experimental social psychologypp. 173–220. New York: Academic Press
Sullenberger, C. B. (2009). Highest Duty: My Search for What Really Matters. New York: Harper Collins Publishers
Miracle on Hudson (2009). Cockpit flight recorder tapes. www.youtube.com/watch#!v=3IIl_AD3Lwo&feature=fvw
Rasskin-Gutman, Diego (2009). Chess Metaphors: Artificial Intelligence and the Human Mind; translated from the Spanish by Klosky, D. Boston: MIT Press
United Nations International Strategy for Disaster Reduction Secretariat (2005): Lessons save lives: the story of Tilly Smith www.preventionweb.net/english/multimedia/v.php?id=970&hid=71