Visual Graphs vs Matrices: Epistemic Limits, Cognitive Preference, and the Design of Understandable Computation

In the paper's words

The ZX calculus and the matrices of linear algebra denote exactly the same maps. A field that knows this has, nonetheless, converged on the pictures. This paper argues that the convergence is not a matter of taste and not a matter of mathematics: it is a fact about bounded cognition.

The abstract, verbatim; every line below is quoted from the text.

What it claims

Mature technical fields converge on the maximally comprehensible representation

Numbers it reports

No sentence of this paper reports a figure with a unit or a quantity word, so none is shown.

How it is done

1. The two-layer structure licenses a design principle, and the principle is larger than quantum computing.

   from 4. The design principle

2. Wen et al., "Quantivine: A Visualization Approach for Large-scale Quantum Circuit Representation and Analysis," arXiv:2307.08969 (2023). [6] M.

   from References

What would falsify it

1. The claim, stated precisely and with falsifiers, is that this two-layer structure is the signature of any mature technical field with hard optimization problems, and that the design principle it licenses is the one the last decade of quantum software has already been practicing: we design systems that design algorithms that optimize themselves and other systems, with legibility as a first-class constraint.

   from Abstract

2. It is stated as a falsifiable hypothesis, with falsifiers, in Section 8.

   from 3. The divergence, and the two-layer structure

3. It is not a metaphor for how things ought to be; it is a description of how the field has already organized itself, and a prediction about where any field with hard optimization problems will end up.

   from 3. The divergence, and the two-layer structure

Where the claim stops

1. A field that knows this has, nonetheless, converged on the pictures: the community's canonical manual is titled Picturing Quantum Software , and the annual conference's technical program is, to a first approximation, representation engineering.

   from Abstract

2. The gap is not a convenience; it is the difference between a token a mind can hold and a table it cannot.

   from 1. The formal equivalence, the cognitive incommensurability

3. Humans cannot optimize; machines cannot explain.

   from 3. The divergence, and the two-layer structure

Where to read

1. Abstract245 words · 1 min
2. 1. The formal equivalence, the cognitive incommensurability161 words · 1 min
3. 2. Why graphs beat matrices for humans283 words · 1 min
4. 3. The divergence, and the two-layer structure220 words · 1 min
5. 4. The design principle132 words · 1 min
6. 5. Evidence anchors from the 2026 conference season213 words · 1 min
7. 6. The tie-in: statistics, thermodynamics, information, re-entry389 words · 2 min
8. 7. Computational verification217 words · 1 min
9. 8. H-VISUAL: the hypothesis card138 words · 1 min
10. 9. Practitioner section196 words · 1 min
11. 10. Where the premises end136 words · 1 min