REVIEW 5 major objections 5 minor 1 cited by
Quantifying the Complexity of Materials with Assembly Theory
T0 review · 5 major / 5 minor · reviewed 2026-08-07 · deepseek-v4-flash
Pith's one-line read The paper argues that Assembly Theory, applied hierarchically to unit cells and their periodic arrangement, measures material complexity and can distinguish engineered from random crystals.
desk verdict The paper's central technosignature claim rests on an asymmetric way of defining internal objects for random vs engineered faults, but the hierarchical-extension work is real and worth refereeing. read the letter →
The pith
A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.
The reading
What carries the argument
The load-bearing object is the nested assembly space of Equation (2), splitting a crystal's assembly index into unit-cell assembly and periodic assembly. It is carried by two quantitative tools: the internal copy number, which counts repeated unit cells or supercells within one material sample, and the Assembly equation $A = \sum_i e^{a_i}(n_i - 1/N_T)$, which exponentiates assembly index so that repeated complex internal objects dominate. For faulted close-packed crystals, the internal object length scale $\xi$ is set by pair-correlation decay for random faults and by fault spacing $\hat{\xi}$ for engineered faults; the hash-assembly algorithm approximates assembly indices on large layered and cellular structures.
What would settle it
Recompute Assembly for the same faulted crystals using one common segmentation length for both random and engineered cases (for instance, both scaled by the pair-correlation length $\xi$); if engineered Assembly no longer exceeds the random maximum of about $10^8$, the claimed technosignature threshold collapses.
Extended reading notes
Core claim
The central claim is that for solid-state materials the assembly space has two nested parts: the shortest path to construct the unit cell from bonds, and the shortest path to construct the object from unit cells, so that $a_{\mathrm{obj}} = a_{\mathrm{uc}} + a_p$. The paper defines an internal copy number for the periodic repetition of unit cells inside a single object and uses the Assembly equation $A = \sum_i e^{a_i}(n_i - 1/N_T)$ to quantify selection. In crystals with stacking faults, the authors compute Assembly from internal objects defined by a correlation length: for random faults, from pair-correlation decay to a 1% threshold; for engineered faults, from regular fault spacing. They report that random faulting yields a maximum Assembly oscillating around $10^8$, while engineered fault patterns can exceed this value, which they interpret as a formalization of a material technosignature.
Load-bearing premise
The comparison of random and engineered crystals assumes they can be segmented at different length scales — random faults by pair-correlation decay to 1%, engineered faults by fault spacing — so the reported separation may depend on that asymmetry rather than an intrinsic material property.
Editorial extensions
If this is right
- Engineered materials can in principle be identified by an Assembly value above the random-fault maximum near $10^8$, giving a measurable technosignature threshold.
- Assembly tracks phase transformations such as HCP-to-FCC and graphite-to-diamond through joint assembly spaces, quantifying loss and discovery of causal constraints.
- Point defects increase the assembly index of a lattice, so real defective materials can be compared using supercell internal copy numbers.
- The trade-off between the number of unique internal objects and their assembly indices provides a handle for optimizing material complexity against function.
Reading between the lines
- A testable extension would apply the same pair-correlation and fault-spacing segmentation to real engineered heterostructures, checking whether the $10^8$ threshold survives with measured copy numbers.
- The asymmetric length-scale choice for random versus engineered faults means the threshold claim should be re-examined under a single shared segmentation scale before being used as a biosignature.
- If the threshold holds, it could turn diffraction or microscopy surveys into screens for technological origin, connecting material complexity to searches for technosignatures.
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The paper extends Assembly Theory (AT) to crystalline and engineered solid-state materials. It proposes that the assembly index of a periodic object decomposes as a_obj = a_uc + a_p (Eq. 2), introduces an 'internal copy number' for repeated unit cells, and applies the framework to mineral databases, point defects, and stacking faults. The central claim is that Assembly computed from internal copy numbers can distinguish engineered from random materials, with engineered stacking faults surpassing a random maximum of about 10^8 (Figure 7E), thereby 'formalizing a material technosignature.' The paper includes a substantial Monte Carlo study and provides code on GitHub.
Significance. If established, the proposed metric would be a meaningful step toward a measurable complexity quantifier for solid-state materials, with potential applications in origin-of-life research, material design, and technosignature detection. The paper contains several useful conceptual contributions: the hierarchical decomposition idea, the internal copy number concept, and the mineral database analysis. The authors also make their code publicly available, which aids reproducibility. However, the central claim about a material technosignature threshold is not yet supported because the comparison between random and engineered systems is uncontrolled, and the foundational additivity assumption (Eq. 2) is not justified. The paper's own admission that precise assembly indices and copy numbers 'do not have experimental meaning' for natural materials further limits the claimed experimental relevance.
major comments (5)
- [Assembly Index of Periodic Crystalline Solids, Eq. (2)] The equation a_obj = a_uc + a_p is introduced as an assumption ('we assume'), but it is used for all subsequent results without proof. For hierarchical assembly, the minimal assembly index of the whole object is not in general equal to the sum of the minimal assembly indices of its parts, because shared substeps or alternative pathways can lower the total cost. The paper should either prove this additivity for the specific assembly spaces considered, or clearly state that a_obj is defined as the sum and discuss the discrepancy from the true minimal assembly index. This is load-bearing because the mineral database analysis and the internal copy number scaling rely on Eq. (2).
- [Assembly Spaces of Hexagonal Closed-Packed Structures with Stacking Faults, Figure 7E] The random and engineered cases are segmented using different rules: for random faults, xi is defined by the pair-correlation decay at a 1% threshold; for engineered faults, xi_hat (the imposed regular fault spacing) is used. Since Assembly in Eq. (1) depends exponentially on the assembly index of the internal objects, and since a larger segmentation length leads to larger assembly indices, the comparison is uncontrolled. The 1% threshold is arbitrary and is applied only to the random case, so the 'random maximum' around 10^8 is a function of that cutoff rather than an intrinsic property. The authors should use the same segmentation rule for both cases or perform a sensitivity analysis over thresholds and demonstrate that the conclusion (engineered > random) is robust.
- [Assembly Spaces of Hexagonal Closed-Packed Structures with Stacking Faults, Figure 7] No error bars or confidence intervals are shown for the Monte Carlo results, despite the text stating that results are averaged over 60 or 10^5 simulations. The central quantitative claim that Assembly 'oscillates around an approximate value of 10^8' requires uncertainty quantification, especially because the technosignature threshold depends on the precise value of this maximum. The authors should plot the distribution or at least provide standard errors for the Assembly values in Figure 7.
- [Distribution of Assembly Indices of Known Minerals] The paper states that it is 'unlikely that it will be possible to determine exact values for unit cells experimentally' and that 'at the bonding level in solid-state materials, the crystallographic analysis averages over an entire crystal,' such that 'precise assembly index and copy number do not have experimental meaning.' This directly contradicts the abstract's claim of experimental complexity measurements and undermines the material technosignature threshold as a measurable observable. The authors should clarify the theoretical status of the threshold and identify what experimental data would be needed to test it, or explicitly reframe the result as a theoretical construct.
- [Periodic unit cells arrangement as an internal copy number] The scaling a ~ a_uc + k log2(n) is stated as a 'first-order approximation' without proof. It is then used to derive A ∝ e^{a_uc} n_k and the internal copy number formalism. For arbitrary periodic objects, the minimal assembly index of the periodic arrangement may not be logarithmic in n, especially when boundaries or defects are present. The paper should either prove this scaling for the specific assembly spaces or discuss the conditions under which it holds; otherwise the internal copy number proportionality is not justified.
minor comments (5)
- [Abstract] The phrase 'allow us formalizing' should be 'allow us to formalize'.
- [Introduction and throughout] 'Hexagonal closed packed' and 'closed-packed' should be 'hexagonal close-packed' and 'close-packed'.
- [Assembly Index of Periodic Crystalline Solids] 'Once the breath-first search is exhausted' should be 'breadth-first search', matching the earlier usage.
- [Figure 3 caption] The notation '4.3k structures' is informal; please write '4,300 structures'.
- [Assembly of Crystalline Structures with Point Defects, Figure 6] The Monte Carlo results in Figure 6B would benefit from error bars or shaded uncertainty regions, consistent with the request in the stacking fault section.
Circularity Check
Figure 7E technosignature threshold is set by asymmetric internal-object segmentation: random faults use a 1% correlation cutoff, engineered faults use their imposed fault spacing, so the predicted separation is partly built into the comparison.
-
fitted input called prediction
[Section 'Assembly Spaces of Hexagonal Closed-Packed Structures with Stacking Faults'; Eq. (1); Figure 7C–E]
"For a closed-packed layer with a given fault probability, the internal object length scale is defined as the length scale such that the correlation probability is less than 1%. ... Next we computed the Assembly values using equation (1) for crystals with correlation lengths ξ̂ ∈ {i ⋅ 0.04% L}i=0 6 as well as for each fault probability up to a fault probability 0.2 L, and we compared these to the example with random faults. In doing this we find that engineered crystals can be more 'complex' than a random crystal."
The Assembly values being compared are not measured under the same object definition. For random faults the internal object size is xi, set by an arbitrary 1% correlation threshold; for engineered faults the internal object size is xi_hat, the imposed regular fault spacing, and xi_hat is swept over a parameter list. In Eq. (1), A = sum e^{a_i}(n_i - 1/N_T), so A is exponentially sensitive to the assembly index of each internal object, and a_i increases with the length of the segment used as the internal object. The 'random maximum' of ~1e8 in Figure 7C is therefore a function of the chosen 1% cutoff, and the engineered curves are generated under a different segmentation rule.
full rationale
Most of the paper is self-contained empirical/algorithmic work: mineral unit-cell assembly indices from AMCSD, the graphite–diamond joint assembly space, and point-defect supercell calculations are computed from explicit structures with stated algorithms and do not reduce to their conclusions. The circularity is localized to the stacking-fault comparison used for the technosignature claim. Assembly in Eq. (1) is computed over 'internal objects' whose length scale is chosen differently in the two arms: random faults are segmented by a 1% pair-correlation cutoff, while engineered faults are segmented by the imposed regular fault spacing, and that spacing is swept. Since A depends exponentially on the assembly index of the segment and the segment index grows with segment length, the relative value of A is strongly controlled by the segmentation convention rather than by an independent material observable. The paper also concedes that exact unit-cell assembly indices are unlikely to be experimentally obtainable, further weakening the framing of this as a measurable technosignature threshold. Hence one central prediction reduces substantially to construction choices; the rest of the derivation stands independently.
Assumptions & free parameters
free parameters (4)
- Internal object length scale threshold =
1% correlation probability
- Regular fault spacing xi_hat sweep =
xi_hat = i * 0.02% * L for i=0..6
- System size L =
L ~ 10,000 layers
- Fault probability beta sweep =
beta values up to 0.25
assumptions (4)
- domain assumption For a hierarchical crystal, the assembly index of the object equals the assembly index of the unit cell plus the assembly index of the periodic arrangement of unit cells.
- ad hoc to paper A crystal can be divided into internal objects of a length scale derived from pair-correlation decay (random) or fault spacing (engineered).
- domain assumption Pair correlation functions for randomly faulted HCP structures decay as P(m) = a +/- b e^{-m/l} with l a function of fault probability beta.
- domain assumption Assembly index can be approximated by the breadth-first or Hash-Assembly algorithms without affecting the qualitative conclusions.
invented entities (2)
-
Internal copy number n_k
-
Internal object length scale xi
Cite this review
Pith. "Pith review of Quantifying the Complexity of Materials with Assembly Theory." pith.science (2026). https://pith.science/paper/INULRHWP
@misc{pith2026250209750,
author = {Pith},
title = {Pith review of: Quantifying the Complexity of Materials with Assembly Theory},
year = {2026},
howpublished = {\url{https://pith.science/paper/INULRHWP}},
note = {Machine review of arXiv:2502.09750}
}
read the original abstract
Quantifying the evolution and complexity of materials is of importance in many areas of science and engineering, where a central open challenge is developing experimental complexity measurements to distinguish random structures from evolved or engineered materials. Assembly Theory (AT) was developed to measure complexity produced by selection, evolution and technology. Here, we extend the fundamentals of AT to quantify complexity in inorganic molecules and solid-state periodic objects such as crystals, minerals and microprocessors, showing how the framework of AT can be used to distinguish naturally formed materials from evolved and engineered ones by quantifying the amount of assembly using the assembly equation defined by AT. We show how tracking the Assembly of repeated structures within a material allows us formalizing the complexity of materials in a manner accessible to measurement. We confirm the physical relevance of our formal approach, by applying it to phase transformations in crystals using the HCP to FCC transformation as a model system. To explore this approach, we introduce random stacking faults in closed-packed systems simplified to one-dimensional strings and demonstrate how Assembly can track the phase transformation. We then compare the Assembly of closed-packed structures with random or engineered faults, demonstrating its utility in distinguishing engineered materials from randomly structured ones. Our results have implications for the study of pre-genetic minerals at the origin of life, optimization of material design in the trade-off between complexity and function, and new approaches to explore material technosignatures which can be unambiguously identified as products of engineered design.
Figures
Figures from the paper (2 more)
Forward citations
Cited by 1 Pith paper
-
Historical Contingencies Steer the Topology of Randomly Assembled Graphs
A recursive graph-merging algorithm produces networks with extreme topological properties that their degree sequences do not predict.
Reference graph
Works this paper leans on
-
[1]
Perspectives in Chemistry—Steps towards Complex Matter
1.Lehn, J. Perspectives in Chemistry—Steps towards Complex Matter . Angew. Chem. Int. Ed. 52, 2836–2850 (2013). 2.Johnston, I. G. et al. Symmetry and simplicity spontaneously emerge from the algorithmic nature of evolution. Proc. Natl. Acad. Sci. 119, e2113883119 (2022). 3.England, J. L. Dissipative adaptation in driven self-assembly. Nat. Nanotechnol. 10...
Reviewed August 7, 2026 · model on record in the stance chip above.
Discussion (0). Continue with ORCID to comment.