A commentary on "From Miller's Seven to Cowan's Four in Baddeley's Working-Memory Framework with Lucas-numbers: Immediate/short-term span L₄ = 7, central working- memory capacity L₃ = 4, and a falsifiable hypothesis based on GEIER's equations - A first approach" by Stefan Geier et al.

A commentary on "From Miller's Seven to Cowan's Four in Baddeley's Working-Memory Framework with Lucas-numbers: Immediate/short-term span L₄ = 7, central working- memory capacity L₃ = 4, and a falsifiable hypothesis based on GEIER's equations - A first approach" by Stefan Geier et al.

From Miller’s Seven to Cowan’s Four: A Conceptually Clarifying Lucas-Number Perspective on Short-Term and Working Memory by Stefan Geier et al.*

This paper is an unusually ambitious and intellectually stimulating contribution at the intersection of experimental psychology, cognitive theory, mathematical pattern analysis and the philosophy of science. Its particular merit lies not merely in observing that the psychologically prominent numbers seven and four are Lucas numbers, but in embedding this observation within a carefully differentiated account of memory constructs. By distinguishing Miller’s classical immediate-memory span of approximately seven chunks from the more restricted capacity of approximately four units associated with controlled working-memory storage, the manuscript corrects a widespread conceptual simplification and replaces it with a more precise theoretical framework.

The paper’s most important achievement is its separation of three historically related but scientifically distinct contributions. Miller’s celebrated “magical number seven, plus or minus two” is appropriately interpreted as a claim concerning immediate memory span, recoding and chunk-based performance rather than as a universal law of working-memory capacity. Baddeley and Hitch are correctly presented as architects of a multicomponent working-memory theory comprising specialized maintenance systems and executive control processes. The frequently cited capacity estimate of approximately four units is then more accurately connected with Cowan’s embedded-processes approach and with experimental paradigms designed to reduce rehearsal, grouping, sensory persistence and long-term-memory support. This historical and theoretical clarification is itself a valuable contribution, independently of the subsequent Lucas-number interpretation.

Within this differentiated framework, the mathematical correspondence is elegant:

$$L_4=7$$

for Miller-like immediate or simple short-term memory span, and

$$L_3=4$$

for the more restricted central capacity commonly observed under controlled working-memory conditions. The proximity of these consecutive Lucas numbers is especially suggestive because the psychological quantities are not treated as interchangeable. Instead, the manuscript proposes that they may describe different functional levels of temporary cognition: a broader, strategy-sensitive span near seven and a more constrained attentional or representational core near four. This is a considerably more sophisticated proposal than simply attaching a preferred number sequence to isolated empirical observations.

The paper also deserves strong praise for its epistemic discipline. It explicitly distinguishes an exact mathematical identity from an empirical regularity and, in turn, from a causal psychological explanation. That $7=L_4$ and $4=L_3$ is mathematically exact. That human memory performance often clusters around these values is an empirical proposition whose validity depends on task design, stimulus domain, age, expertise, rehearsal opportunities, chunking strategies and the operational definition of capacity. Whether Lucas-number structure has any deeper explanatory or causal role remains a hypothesis requiring independent testing. By maintaining these distinctions, the manuscript avoids the common error of treating numerical correspondence as automatic proof of mechanism.

This truth-oriented structure substantially strengthens the discussion of GEIER’s equations and their proposed Fibonacci–Lucas extensions. The paper does not present the psychological observations as already established confirmation of a universal mathematical programme. Rather, it identifies them as potentially relevant test cases within that programme. Such restraint is scientifically productive: it transforms a striking numerical observation into a falsifiable research question. The paper thereby moves the discussion away from retrospective pattern matching and toward prospective experimentation, model comparison and preregistered prediction.

A further major strength is the proposed experimental programme. The manuscript recognizes that an adequate test cannot consist merely of asking whether sample means happen to lie near four or seven. It therefore calls for paradigms that manipulate rehearsal, recoding, attentional control, grouping and representational complexity. This makes it possible to examine whether performance shifts systematically between a broader $L_4=7$ regime and a more constrained $L_3=4$ regime. The inclusion of competing models—such as fixed-slot accounts, variable-precision models, continuous-resource theories and unrestricted discrete-state models—is especially commendable. A Lucas-based hypothesis becomes scientifically informative only when it can outperform plausible alternatives or make distinctive predictions that those alternatives do not make.

The manuscript is also commendable for treating four and seven as approximate psychological attractors rather than inflexible biological constants. Human cognitive performance is inherently variable, and experimental estimates depend heavily on the measurement model. The paper therefore wisely avoids claiming that every individual, task or sensory modality must produce an integer capacity of exactly four or seven. Instead, it asks whether these values occupy privileged positions in the organization of temporary memory once relevant confounds are controlled. This formulation is both more realistic and more experimentally tractable.

Conceptually, the paper opens an intriguing possibility: the celebrated discrepancy between the “magical number seven” and the later “magical number four” may not represent a simple historical correction in which one number replaces the other. Both values may be valid within different cognitive regimes. Seven may characterize immediate span when chunking, rehearsal and learned coding structures contribute substantially, whereas four may better characterize the limited central workspace available when those supports are minimized. The Lucas sequence provides a concise formal language for expressing this layered relationship without erasing the psychological differences between the constructs.

The paper is therefore valuable on several levels. Historically, it clarifies the distinct meanings of Miller’s, Baddeley’s and Cowan’s contributions. Psychologically, it proposes a structured distinction between simple short-term span and controlled working-memory capacity. Mathematically, it identifies the exact relations $L_4=7$ and $L_3=4$. Philosophically, it differentiates formal identity, empirical fit, explanatory hypothesis and causal mechanism. Methodologically, it outlines how the proposal could be tested rather than merely admired.

Overall, this is a creative, carefully qualified and potentially generative paper. Its strongest feature is the combination of intellectual boldness with methodological caution. The Lucas-number correspondence is presented neither as a trivial curiosity nor as a prematurely established law of cognition. It is developed as a formally precise observation that motivates a new experimental research programme. By showing that Miller’s seven and the controlled working-memory estimate of four can be interpreted as consecutive Lucas numbers while retaining their distinct psychological meanings, the paper offers an original integrative perspective on one of the most enduring numerical motifs in cognitive psychology.

MGN

*Geier Stefan et al.: From Miller's Seven to Cowan's Four in Baddeley's Working-Memory Framework with Lucas-numbers: Immediate/short-term span L₄ = 7, central working- memory capacity L₃ = 4, and a falsifiable hypothesis based on GEIER's equations - A first approach. ResearchGate, July 2026, DOI:
10.13140/RG.2.2.12182.72003 .