Cocycle Solving in Cognition: Ultrametric Reaction Times and Analogy Depth

In the paper's words

Cognitive consistency and analogical difficulty are pervasive challenges for psychologists, educators, and artificial‑intelligence designers. Existing Euclidean models cannot capture the hierarchical constraints that appear in belief networks and analogy spaces. We therefore propose a formalism in which the brain enforces a cocycle (flatness) condition on an associative graph, yielding a globally ultrametric semantic topology.

The abstract, verbatim, and its opening sentence is the claim. Every line below is quoted from the text.

Numbers it reports

1. Using the strong triangle inequality from ultrametric theory ([1]), the hierarchical processing view of thalamocortical loops ([2]), and the embodied complexity framework ([3]), we compute concrete RT values (e.g., a depth‑4 stimulus predicts 400 ms) and difficulty scores (e.g., a cross‑ratio of 9 predicts $\approx 220$ ms).

   from Abstract

2. The RT prediction is linear in $b$; a 10 % increase in $b$ raises $RT(4)$ by $8\,$ms.

   from 4.4 Sensitivity to Parameters

3. For a depth‑4 semantic pair, the model predicts a reaction time of 280 ms .

   from 5. Results

4. Numerical examples: depth‑4 stimulus predicts $RT=280\,$ms; cross‑ratio = 9 predicts $D\approx220\,$ms (Section 5).

   from 1. Introduction

5. Assume ultrametric distances (in abstract depth units) for an analogy problem: $$ d_u(A,C)=3,\quad d_u(B,D)=3,\quad d_u(A,D)=1,\quad d_u(B,C)=1. $$ Compute the cross‑ratio: $$ \begin{aligned} \mathrm{CR} &= \frac{d_u(A,C)\,d_u(B,D)}{d_u(A,D)\,d_u(B,C)} \\ &= \frac{3\times 3}{1\times 1} \\ &= 9. \end{aligned} $$ Now compute difficulty: $$ \begin{aligned} D &= k\,\log(\mathrm{CR}) \\ &= 100\;\text{ms}\times \log(9) \\ &= 100\;\text{ms}\times 2.19722458\;(\text{natural log}) \\ &\approx 219.722\;\text{ms}. \end{aligned} $$ Rounded to the nearest millisecond, $D\approx220\,$ms.

   from 4.3 Analogy‑Depth Computation

6. For an analogy with cross‑ratio = 9, the model predicts a difficulty score of $\approx 220$ ms .

   from 5. Results

How it is done

1. From this topology we derive explicit reaction‑time (RT) predictions that grow linearly with the ultrametric depth of a stimulus, and we formalize “analogy depth’’ as the projective cross‑ratio of the four terms in an A : B :: C : D problem.

   from Abstract

2. Given $\phi$, we sort concepts by their potential values and construct a rooted tree $T$ where each leaf corresponds to a concept.

   from 3.2 Construction of the Ultrametric Tree

3. Empirical work on semantic priming suggests that RT grows with the hierarchical distance between prime and target.

   from 3.3 Reaction‑Time Model

4. The projective cross‑ratio is $$ \mathrm{CR}= \frac{d_u(A,C)\,d_u(B,D)}{d_u(A,D)\,d_u(B,C)}. $$ We define analogy difficulty as $$ D = k\,\log(\mathrm{CR}), $$ with $k$ a scaling constant converting logarithmic distance into milliseconds.

   from 3.4 Analogy Depth via Cross‑Ratio

What would falsify it

1. A single experiment that measures RTs for a set of semantic priming pairs spanning depths $L=1$ to $L=5$ and finds a statistically significant deviation from the linear relationship $RT = a + bL$ (e.g., curvature or plateau) would falsify the core RT prediction.

   from 6.3 Falsification Criterion

2. Similarly, if difficulty ratings for analogies do not correlate with $\log(\mathrm{CR})$ across a diverse stimulus set, the analogy‑depth claim would be falsified.

   from 6.3 Falsification Criterion

3. Using the strong triangle inequality from ultrametric theory ([1]), the hierarchical processing view of thalamocortical loops ([2]), and the embodied complexity framework ([3]), we compute concrete RT values (e.g., a depth‑4 stimulus predicts 400 ms) and difficulty scores (e.g., a cross‑ratio of 9 predicts $\approx 220$ ms).

   from Abstract

Where the claim stops

1. Assumption Sensitivity Baseline $a=200\,$ms Low: changes affect all RTs uniformly.

   from 6.1 Limitations

2. Future work should empirically test the depth‑RT relationship and explore neural implementations of the cocycle condition, thereby advancing a hard geometric foundation for cognition.

   from 8. Conclusion

3. Existing Euclidean models cannot capture the hierarchical constraints that appear in belief networks and analogy spaces.

   from Abstract

Where to read

1. Abstract203 words · 1 min
2. 1. Introduction288 words · 1 min
3. 2. Background and Related Work375 words · 2 min
4. 3. Methods295 words · 1 min
5. 4. Analysis209 words · 1 min
6. 5. Results68 words · 1 min
7. 6. Discussion211 words · 1 min
8. 7. Practical Implications and Engineering Pathways256 words · 1 min
9. 8. Conclusion84 words · 1 min
10. 9. Extended Literature Synthesis281 words · 1 min
11. Correction Notice305 words · 1 min
12. Provenance and AI Involvement238 words · 1 min