Hierarchical Complexity

From: jimruttshow8596

Introduction

Hierarchical complexity is a concept primarily explored within developmental psychology and psychometrics, focusing on the qualitative differences in task difficulty and cognitive development [02:08:08]. It is a fundamental property of mental processes [04:25:05] and the construction of skill [08:13:20].

Horizontal vs. Vertical (Hierarchical) Complexity

The concept of hierarchical complexity is often contrasted with horizontal complexity [02:22:16].

• Horizontal Complexity: Refers to doing more of the same task at the same level of complexity [06:36:03]. For example, tying a thousand shoelaces [07:14:02] or navigating an elevator by pressing a series of disjoint buttons [11:30:29]. These tasks might overwhelm capacity due to quantity or distractions [07:42:15].
• Hierarchical (Vertical) Complexity: Involves performing a fundamentally qualitatively more complex task [06:47:33]. This demands the integration of lower-order capacities that have already been mastered [07:54:02]. For instance, making a shoe is hierarchically much more complex than tying a shoelace [07:07:23]. Another example is disassembling, replacing a part, and reassembling a lawnmower engine, where every move depends on others and requires ancillary skills like organizing parts [11:55:04].

Ultimately, any skill contains some level of hierarchical complexity [12:32:00]. Even simple acts like focusing on an elevator button or reaching and pushing it involve complex coordination of sensory motor skills that young children cannot perform [12:51:30].

Origins and Key Figures

The construct of hierarchical complexity has a long research history in psychology, with its origins dating back almost a century to Jean Piaget’s 1923 publication, The Language and Thought of the Child [04:02:18]. Earlier shades of the concept can be found in thinkers like James Mark Baldwin and Herbert Spencer [04:15:20].

In the 1980s and 1990s, a “neo-Piagetian consensus” emerged on the formal definition of developmental stages [04:32:00]. Key figures in this development include:

• Jean Piaget: His work on the semiotic function, where sprawling sensory motor experiences are summarized into a single representation (e.g., “bedtime”), was seen as miraculous and a key differentiator of humans from animals [14:55:09].
• Michael Commons: Coined the term “Model of Hierarchical Complexity” [03:38:00] and articulated it almost as a mathematical construct [09:47:00].
• Kurt Fischer: A prominent neo-Piagetian who conducted empirical work to validate HC, emphasizing its manifestation in human skill and behavior (Dynamic Skill Theory) [03:23:00] [09:40:00]. Fischer also studied HC in animals like monkeys and pigeons [19:18:00].
• Theo Dawson: Pioneered psychometric innovation around the construct of HC, creating the “Lectical Levels” measurement system [03:15:00] [05:59:00].

Hierarchical Integration and “Chunking”

Hierarchical complexity is characterized by “hierarchical integration,” where many lower-order skills become a single, qualitatively new, more complex skill [08:57:00]. This process means that once lower-order elements are integrated into a higher-order element, they are no longer the same elements they were before [00:37:34].

This process is deeply related to the concept of “chunking” [16:08:00]. Chunking refers to loading more meaning and depth into a single term or concept as experience grows [16:37:00]. For example, a veteran complexity scientist understands the word “complexity” with a “huge history of depth and nuance” compared to a layperson [16:51:00].

Hierarchical complexity proposes a definable scale of chunking [17:33:00]:

• Chunking sensory motor systems.
• Chunking theories of biological evolution into higher-order theories [17:40:00].

Chunking is a ubiquitous property, extending down to reflexes and nervous system action, as seen in the eye’s distillation of signal noise comparable to chunking [18:03:00]. Early psychologists like William James and Charles Sanders Peirce observed this as a basic property of the nervous system [18:20:00].

Hierarchical Complexity as an Ontological Property

The property of hierarchical complexity, as articulated by Michael Commons, is described as almost a general property of information across the biological spectrum [19:02:00]. It is seen as a continuous evolutionary process where higher-order or hierarchical complexity emerges [21:42:00].

This connects to the broader science of complexity, which is fundamentally about the emergence of complexity from simplicity over multiple levels [19:54:00]. A human being is a prime example: atoms combine into molecules, then long chains, then metabolism in cells, multicellular organisms, organs, and organ systems, eventually leading to entities living in ecosystems – all stemming from simple atoms [20:01:00]. This is a characteristic attribute of complex adaptive systems that exhibit emergence from the bottom up [21:04:00].

Psychology, particularly developmental psychology, studies how the individual human evolves as an instantiation of an organism that evolved out of an evolving universe [22:30:00]. Piaget’s insight was that epistemology and psychology must integrate with Darwinian evolution [22:43:00]. The fact that nature can be asked “ever more complex questions” and that the universe is lawful suggests an implicit realism behind the drive for hierarchical complexity [31:30:00] [31:38:00]. If patterns were truly random, investing in the biological cost of extracting abstract patterns would be useless [31:45:00].

Drivers of Hierarchical Complexity

Several factors drive the emergence of hierarchical complexity in the human mind:

• Working Memory Bottleneck: The remarkably small capacity of short-term memory (e.g., Miller’s 7 ± 2 for sound, 3-4 for images) is a significant driver [27:27:00]. To solve more difficult problems, the mind must make chunks larger, enabling more information to be processed simultaneously [27:57:00].
• Demands of the World: The inherent complexity of the environment is a primary driver [29:31:00]. If the world were not complex, a working memory bottleneck would not necessitate higher-order organization [29:33:00]. Conversely, a lack of complexity or consistency in the environment can hinder the nervous system’s expectation and pursuit of complexity [32:03:00].
• Epistemic Motivation (Need to Understand): Humans possess an innate desire to understand, not just to succeed [29:50:00]. Children eventually cross a threshold from wanting success to wanting to know why something worked [30:06:00]. This motivation propels information through the memory bottleneck [30:41:00].

Measurement and Levels

Measuring complexity in human cognitive development is typically done by looking at open-ended human performance, often linguistic, to discern the underlying hierarchical structure [01:13:24].

Theo Dawson’s definition of hierarchical complexity emphasizes:

“Hierarchical complexity refers to the number of non-repeating recursions that coordinating actions must perform on a set of primary elements. Actions at the higher order of hierarchical complexity are defined in terms of the actions at the next lower level. A and B organize and transform the lower order actions and C produce organizations of lower order actions that are new and not arbitrary and cannot be accomplished by the lower orders themselves.” [35:28:00]

This formal, mathematical definition allows for its application in areas like computer programming and information processing systems [38:19:00].

The Commons model specifies 15 levels of hierarchical complexity [39:05:00]. Kurt Fischer’s system uses tiers and levels, which are almost identical to Commons’s model and Dawson’s Lectical Levels [41:11:11]. These tiers are:

1. Actions (Sensory Motor Tier): Begins with single actions (e.g., focusing on a face) [42:58:00], progressing to coordination of actions (e.g., looking and reaching) [43:08:00], and eventually systems of actions (e.g., coordinating looking, reaching, grasping, and drinking) [43:17:00]. Development in this tier is rapid in infancy, occurring in weeks or days [44:44:00].
2. Representations (Operational/Primary Concrete Tier): Emerges as a child’s “head pops up” to a new world of using linguistic signs to represent non-present realities [47:52:00]. This includes entertaining counterfactuals, lying, or talking about imaginary figures like Santa Claus [48:19:00]. It begins with single representations (e.g., “mommy,” “doggy,” “bedtime”) [48:32:00], then mapping them together (e.g., “mommy water”) [48:40:00], and eventually developing representational systems that allow for complex descriptive stories [49:09:02].
3. Abstractions (Formal Abstract/Systemic Tier): Where many examples of representations are integrated into higher-order concepts [50:22:00]. Unlike representations that can be simply pointed to, abstractions integrate multiple representations (e.g., “quality family time” generalizes across “bedtime,” “dinner time,” “TV time”) [50:36:00]. Most essential aspects of civilization exist at this level [51:00:00], and achieving this level requires education, being crucial for concepts like democracy and justifying law [51:09:00].
4. Principles (Meta-Systemic Tier): Involves a multi-systemic integration of several different fields into a single theory [01:09:32]. An example in moral judgment is John Rawls’s justice as fairness, which is a principled definition of fairness that transcends and includes concrete and abstract notions of fairness [01:09:38]. This level allows for “norming the norms of the convention and making rules about the making of rules” [01:10:37].

Hierarchical complexity also exhibits a fractal-like property: the closer one looks, the more levels and sub-levels of skill construction can be observed [42:38:00].

Applications

Educational Contexts

Hierarchical complexity can be operationalized in educational contexts [09:36:00]. The Lectica assessment system, founded by Zach Stein and Theo Dawson, is a standardized diagnostic tool based on HC [01:13:57]. Unlike traditional standardized tests that classify a student as pass/fail, Lectica assessments provide diagnostic information, identifying what a student understands and what would be the next best thing for them to learn [01:15:46]. This is possible because the system is built upon empirically grounded “rational reconstructions” of specific learning sequences [01:15:59]. This allows for tailoring learning experiences to an individual’s “zone of proximal development” [01:16:44].

Business and Leadership Development

In business and government, hierarchical complexity assessments are used to identify and grow leaders [01:18:10] [01:17:57]. The demands on leaders, particularly in large organizations like intelligence agencies, are qualitatively more complex than in the past [01:19:01].

Research suggests a “complexity gap” between the demands of most leadership roles and the capacities of leaders [01:19:19]. A key area where struggles emerge is in perspective-taking, perspective-seeking, and perspective-integration [01:19:54]. While some leaders may be able to imagine others’ perspectives, they may not habitually seek them out [01:22:24]. These skills are distinct; one can be good at taking perspectives without seeking or integrating them, but effective integration requires mastery of the other two [01:22:09].

The complexity of an organization’s structure can also impact perspective-taking. Flattening organizations and increasing the number of direct reports for managers (e.g., from 5-7 to 20) can increase the difficulty of a manager’s perspective-taking task [01:28:50]. In environments with non-democratic or zero-sum competitive dynamics, individuals may be disincentivized from genuine perspective-taking [01:30:07].

Leadership assessments based on HC are designed to promote development, not just for hiring or firing [01:24:06]. They offer diagnostic reports that reveal areas for growth and provide educational support [01:23:58]. These measures are considered more “ecologically valid” than personality tests like Myers-Briggs or IQ tests, as they directly assess how an individual makes decisions on the job [01:25:20].

Hierarchical Complexity vs. General Intelligence

While there are correlations between hierarchical complexity and measures of general intelligence (g-factor or IQ tests), they are not the same [00:57:54].

• IQ tests often act as a “summary statistic” based on a narrow range of indices [00:59:02]. They tend to categorize an “entire person” as smarter than another, rather than specifying skill in particular domains [00:59:30]. This approach is rooted in eugenics and oversimplifies human psychology, creating a “demi-reality” [01:00:08].
• Hierarchical complexity does not assign a complexity level to a person. Instead, it assigns a level to a particular task they accomplished [00:59:15]. This distinction acknowledges that an individual can demonstrate highly complex reasoning in one domain (e.g., physics) but perform at a lower complexity level in another (e.g., small engine repair) [01:01:29].

Cautions and Misapplications

A significant caution is against the misapplication of hierarchical complexity by “simple stage thinkers” or those who use “color-coded developmental language” to classify entire populations or individuals [01:03:54]. For example, some misuse HC to suggest that “indigenous people were like children” or to apply the individual developmental story directly to the history of civilization [01:00:08]. This oversimplification is problematic because:

• There hasn’t been enough empirical research on HC to reliably generalize about large populations [01:03:21].
• It revives the dangers of social sorting mechanisms seen with IQ tests [01:05:51].
• It can become a “bludgeon” used to categorize and judge people, which is contrary to its intended purpose of promoting growth and understanding [01:05:54].
• It risks creating a “darth vader” scenario, where highly complex individuals use their advanced perspective-taking abilities for strategic or malevolent reasons [01:32:26].

The focus should be on understanding the “consistency of performance within a developmental range” for a single person, rather than classifying entire individuals or populations at a fixed level [01:05:27].