🧬 biology

Introducing Lithikos 3 – The Solid Geometry of Latent Acheulean Knapping Expertise: A Novel Referenced Volumetric Morphometric Approach

This paper introduces Lithikos 3, a novel sample-independent volumetric morphometric approach that quantifies latent Acheulean knapping skill through a persistent scalar coefficient derived from an artefact's solid geometry, thereby extending analytical capabilities from assemblage-level tendencies to individual artefact-level distributional structures.

Original authors: James Hicks

Published 2026-07-23

📖 6 min read🧠 Deep dive

CC BY 4.0

Original authors: James Hicks

Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). ⚕️ This is an AI-generated explanation of a preprint that has not been peer-reviewed. It is not medical advice. Do not make health decisions based on this content. Read full disclaimer

Imagine you are a detective trying to solve a mystery that happened 300,000 years ago. But instead of fingerprints or footprints, your only clues are rocks. Specifically, you are looking at ancient stone tools called handaxes, which early humans chipped away from larger stones to create sharp, teardrop-shaped weapons. For decades, scientists have been obsessed with these tools, not just because they are cool, but because they are a window into the minds of our ancestors. If a stone tool is perfectly symmetrical and incredibly thin, it suggests the maker had a high level of skill, patience, and perhaps even a "mental template" in their head before they started hitting the rock.

The big question has always been: How do we measure that skill? For a long time, scientists used rulers and calipers to measure the distance between points on the stone. They calculated ratios, like "how long is it compared to how wide?" and "how flat is it compared to how thick?" It was like trying to understand a 3D sculpture by only measuring the length of its shadow. The problem is that these measurements depend entirely on which group of rocks you happen to be looking at. If you add a new rock to your pile, the "average" skill level of the whole group might shift, making it impossible to say if one specific rock is truly "better" than another without re-measuring everything. It's like trying to judge a runner's speed by comparing them to the other people in the race; if you swap the other runners for faster ones, your runner suddenly looks slower, even though their actual speed hasn't changed.

This is where a new study by James Hicks from the University of Oxford comes in. He introduces a fresh way to look at these ancient tools, moving away from simple ruler measurements to a method that treats the stone as a solid, 3D object. He calls his new tool "Lithikos 3." Instead of just measuring lines, Lithikos calculates the actual volume and shape of the stone in a way that is independent of the other rocks in the pile. It's like giving every single handaxe its own permanent "skill score" that stays the same no matter who else is in the room.

The New Method: A 3D "Skill Score"

The paper argues that making a handaxe is a three-dimensional act of moving mass around. A skilled knapper doesn't just make a flat shape; they carefully redistribute the weight of the stone to make it thin, symmetrical, and long, all while keeping it strong enough not to break. Hicks suggests that to truly measure this skill, we need a method that respects the 3D nature of the stone.

He developed a mathematical formula that combines two main things: Refinement (how thin and flat the tool is relative to its size) and Symmetry (how perfectly the left side mirrors the right side, and the top mirrors the bottom). By using high-resolution 3D scans of the stones, his software creates a "perfectly symmetrical" ghost version of each tool and compares the real tool to it. It then crunches the numbers to produce a single, persistent number called the Knapping Expertise (KE) coefficient.

Think of it like a video game character's stats. In the old way of doing things, a character's "strength" was only known relative to the other characters in the current level. If you added a super-strong boss to the level, everyone else looked weaker. Hicks' new method gives every character a fixed strength stat that never changes, regardless of who else is in the game. This allows scientists to compare a tool from one ancient site directly with a tool from a completely different site, thousands of miles away, without needing to re-calculate the whole group.

What the Study Found

To test this new method, Hicks analyzed 587 real handaxes from six different sites in the Levant (an area in the Middle East including parts of modern-day Israel and Jordan). These tools ranged from 0.750 million years ago to 0.200 million years ago. He also tested the method on 63 replica handaxes made by a modern expert knapper named Will Attard to see if the math worked on tools known to be made with extreme skill.

The results were fascinating. The study found that the new "skill score" (KE) revealed patterns that the old ruler-based methods missed.

Different Paths to the Same Goal: The study suggests that there isn't just one "perfect" way to make a skilled handaxe. Some ancient groups focused on making their tools incredibly thin and long (high refinement), while others focused more on making them perfectly symmetrical. Both approaches resulted in high skill scores. It's like two different chefs making a great cake: one uses perfect measurements, and the other uses perfect timing; both end up with a delicious result, but their methods look different.
Surprising Ancient Masters: The analysis showed that the site of Nahal Zihor B (NZB), which is from the Late Acheulean period, had the highest average skill scores. However, the site of Gesher Benot Ya'aqov (GBY), which is much older (Middle Acheulean), also showed evidence of incredibly high skill, particularly in making thin, elongated tools. This suggests that the ability to make these sophisticated tools appeared earlier in human history than some previous theories suggested.
The "Twisted" Exception: The study also looked at a special type of handaxe called a "twisted ovate," which has an S-shaped edge and is often considered the most beautiful and difficult to make. Surprisingly, the new skill score didn't rate these as the absolute highest. The expert knapper who made the replicas explained that while twisted tools are tricky, making a very thin, straight, and long tool actually requires more precise control over the stone's volume. The new math reflects this: it rewards the structural control needed to make a tool thin and long, rather than just the visual "wow" factor of a twist.

What This Means

The paper doesn't claim to have solved the mystery of human evolution, but it offers a much sharper tool for looking at it. By moving from "relative" measurements (how does this rock compare to that one?) to "intrinsic" measurements (what is the actual skill level of this specific rock?), scientists can now build a cumulative database. They can add new discoveries from different parts of the world and different time periods, and the numbers will still make sense together.

The study suggests that the "variable sameness" of these ancient tools—where they all look similar but have subtle differences—isn't just random noise. It represents a flexible system where skilled humans found different geometric solutions to the same problem. Some focused on thinness, others on symmetry, but they all achieved a high level of expertise. The new method, Lithikos 3, allows us to see these different paths clearly, proving that the ancient stone knappers were not just following a rigid rulebook, but were creative engineers mastering the geometry of their world.

Technical Summary: Introducing Lithikos 3 – The Solid Geometry of Latent Acheulean Knapping Expertise

Problem Statement
Current methodologies for quantifying latent stone knapping expertise in Acheulean handaxes rely heavily on distance-derived, relational approaches (DD-RAs). These methods typically employ multivariate statistical analyses, such as Principal Component Analysis (PCA), on linear metrics (e.g., Euclidean distances, ratios of length, breadth, and thickness) or unreferenced volumetric measures. The author identifies three primary limitations in this framework:

Sample-Dependency: Results are contingent upon the specific covariance structure of the analyzed dataset. An individual artifact's position in the morphospace is relative and shifts if the sample composition changes, preventing direct comparison across different assemblages or studies without re-analysis.
Ephemerality: The construct of "knapping skill" is not a stable property of the artifact itself but a fleeting statistical abstraction derived from the group.
Dimensional Abstraction: Linear metrics and ratios often fail to capture the full three-dimensional solid geometry of the tool. They quantify distances between discrete points or mass distribution across arbitrary planes without being dimensionally commensurate with the volumetric nature of the object. Furthermore, existing volumetric measures often lack a unified reference frame, conflating 2D and 3D geometric frameworks.

Methodology
The paper introduces Lithikos 3, a novel referenced volumetric morphometric approach designed to quantify knapping expertise as an intrinsic, persistent property of individual artifacts. The methodology centers on the Knapping Expertise (KE) coefficient, a scalar value derived from the high-resolution solid geometry of the handaxe.

The KE coefficient is calculated through the following steps:

Radial Symmetry ( $\omega$ ): The method generates two custom, radially symmetrical reference models for each artifact: an internal reference (Endo-Sym) and an external reference (Exo-Sym). These are synthesized based on the artifact's principal axis (Tip-to-Center of Mass). The Omega Coefficient ( $\omega$ ) is the arithmetic mean of the volumetric ratios between the artifact and these symmetrical references, quantifying how evenly mass is redistributed around the primary axis.
Size-Standardized Refinement ( $VRI_{100k}$ ): The Volumetric Refinement Index (VRI) is calculated using the artifact's volume ( $V$ ), surface area ($SA$), breadth ( $B$ ), and length ( $L$ ). To control for allometric effects (where smaller tools might appear artificially refined), the raw VRI is scaled to a common target volume of 100,000 mm³, resulting in $VRI_{100k}$ .
KE Calculation: The final KE coefficient is defined as a function of these two predictors: $KE = 10\omega - VRI_{100k}$ . This formula integrates radial mirror symmetry and size-standardized bifacial thinness (refinement) into a single, dimensionally commensurate scalar.

The study applied this methodology to two datasets:

Archaeological Sample: 587 handaxes from six Levantine Middle and Late Acheulean sites (Gesher Benot Ya'aqov, Jaljulia, Nahal Zihor B, Revadim, Ma'ayan Barukh, and Holon).
Experimental Sample: 63 high-resolution 3D models of replica handaxes knapped by an expert (Will Attard) to verify the coefficient's behavior at the upper range of skill.

Key Results

Artifact-Intrinsic Stability: The KE coefficient successfully assigns a persistent value to individual handaxes independent of sample composition. This allows for direct comparison of knapping expertise between artifacts from different sites and chronological periods without reconstructing a shared morphospace.
Distributional Analysis: Unlike DD-RAs which focus on central tendencies, the KE framework enables the analysis of expertise distributions (variance, spread, and upper-tail behavior). The study found that while sites differed in median KE values, their variances were not statistically significantly different, suggesting overlapping ranges of skill within communities.
Configurational Pathways: The analysis revealed that comparable levels of expertise (KE scores) can be achieved through different geometric configurations. Some assemblages achieved high KE through exceptional refinement and elongation with moderate symmetry, while others relied more heavily on symmetry. This supports the hypothesis that knapping expertise is a configurational property rather than a linear progression toward a single ideal form.
Site-Specific Findings:
- Nahal Zihor B (NZB) exhibited the highest median KE and upper-tail values, indicating exceptional refinement and elongation in the Late Acheulean.
- Gesher Benot Ya'aqov (GBY), a Middle Acheulean site, also displayed significantly higher expertise than other Late Acheulean sites, challenging narratives of a strictly linear improvement in skill over time.
- Twisted Ovates: In the experimental sample, "twisted ovate" handaxes (often considered aesthetically complex) received moderate KE scores. The study notes that while these forms require specific procedural sequences, they do not necessarily score higher on the KE metric than thin, elongated, highly symmetrical tools, as the twisting reduces the radial symmetry component of the coefficient.

Significance and Claims
The paper claims that the KE coefficient represents a methodological shift from relational, distance-based morphometry to intrinsic, volumetric morphometry. Its primary significance lies in:

Dimensional Commensurability: By using volumetric units to measure a volumetric object, the method aligns the metric with the ontology of the artifact, avoiding the dimensional abstraction inherent in linear ratios.
Portability and Cumulative Science: Because KE values are sample-independent, they facilitate the creation of cumulative databases. New artifacts can be added to existing analyses without recalculating the values of previously analyzed items, enabling longitudinal and cross-regional meta-analyses.
Theoretical Alignment: The approach aligns with Material Engagement Theory (MET) and 4E cognition frameworks by treating the artifact's geometry not just as a passive outcome of skill, but as an active, co-constitutive component of the cognitive system. It quantifies the "solid geometry of latent expertise" as a stable property of the material form.
Complementarity: The author explicitly states that KE does not render distance-based approaches obsolete. Instead, the two are complementary: relational methods (DD-RAs) are effective for identifying covariance structures and clustering within specific datasets, while KE provides a stable, intrinsic measure of performance that allows for cross-dataset comparison and distributional analysis of skill.

The study concludes that knapping expertise is best understood as a flexible, multi-configurational property that emerges from the interaction of symmetry and refinement, rather than a unilinear trajectory toward an idealized morphology.

More like this