arxiv: 2605.00778 · v1 · submitted 2026-05-01 · 💻 cs.LG · q-bio.NC

Recognition: unknown

Observable Performance Does Not Fully Reflect System Organization: A Multi-Level Analysis of Gait Dynamics Under Occlusal Constraint

Jacques Raynal , Pierre Slangen , Jacques Margerit

Authors on Pith no claims yet

Pith reviewed 2026-05-09 20:11 UTC · model grok-4.3

classification 💻 cs.LG q-bio.NC

keywords gait dynamicsocclusal constraintParkinson's diseasestate space analysislatent space embeddingmulti-level analysisneuromechanical systemadaptive systems

0 comments

The pith

Conditions with matching gait performance can stem from distinct underlying organizations in state space and latent embeddings.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper tests whether observable walking performance reliably indicates how a neuromechanical system is organized internally. It imposes controlled changes in jaw height as a constraint on a single patient with Parkinson's disease and measures gait across repeated trials. Standard performance numbers such as speed or step timing are compared against movement trajectories plotted in state space and against clusters found by unsupervised embedding methods. The analysis finds that some conditions produce nearly identical performance scores yet occupy different regions and show different temporal patterns in the trajectory and embedding spaces. This matters because it shows that relying solely on output metrics can overlook real differences in how the system adapts to the same constraint.

Core claim

Conditions with comparable observable performance may correspond to different organizations in both state space and latent space representations. This dissociation highlights a limitation of aggregated metrics and suggests that similar outputs may arise from non-equivalent system states. The study applies the vertical dimension of occlusion as a constraint on gait in one Parkinson's patient, analyzes the data at three levels (aggregated linear metrics, dynamical state-space trajectories, and unsupervised latent embeddings), and concludes that the levels do not always align. A purely conceptual fourth level describing potential state relationships is outlined but not implemented with data.

What carries the argument

A three-level analysis framework that compares aggregated performance metrics against state-space trajectory organization and unsupervised latent embeddings of gait time series under varying occlusal constraints.

If this is right

Aggregated performance metrics alone cannot distinguish between non-equivalent system organizations under the same constraint.
State-space and latent-space representations can reveal organizational differences invisible at the performance level.
Unsupervised embedding provides a data-driven way to detect these differences without requiring labeled classes.
The multi-level structure supplies a systematic approach for examining how adaptive systems respond to imposed constraints.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

Clinical gait evaluations in neurological conditions could incorporate trajectory and embedding views to detect adaptations that performance scores miss.
The same performance-organization mismatch may occur in other motor behaviors constrained by dental or postural changes.
Extending the analysis to multiple participants would test whether the dissociation generalizes beyond the single case examined here.

Load-bearing premise

Observed differences in state-space trajectories and latent embeddings reflect genuine variations in neuromechanical organization rather than arising from the specific variables chosen, embedding settings, or single-subject variability.

What would settle it

If additional trials under identical occlusion conditions repeatedly produced overlapping state-space trajectories and identical latent embeddings whenever the aggregated performance metrics matched, the claimed dissociation would not hold.

Figures

Figures reproduced from arXiv: 2605.00778 by Jacques Margerit, Jacques Raynal, Pierre Slangen.

**Figure 4.** Figure 4: Conceptual illustration of potential relationships between system states. Visually distinct [PITH_FULL_IMAGE:figures/full_fig_p011_4.png] view at source ↗

read the original abstract

In biomechanical systems, observable performance is often used as a proxy for underlying system organization. However, this assumption implicitly presumes a correspondence between output metrics and internal system states that may not hold in adaptive systems. In this study, the vertical dimension of occlusion (VDO) is considered as a constraint applied to an adaptive neuromechanical system, enabling the exploration of system-level responses under controlled variations. A single-case design in a patient with Parkinson's disease allows an intra-individual analysis across repeated conditions.The analysis is structured across three complementary levels: (i) aggregated linear metrics describing observable performance, (ii) a dynamical systems framework describing temporal organization in state space, and (iii) a latent space representation obtained through unsupervised embedding. The results show that conditions with comparable observable performance may correspond to different organizations in both state space and latent space representations. This dissociation highlights a limitation of aggregated metrics and suggests that similar outputs may arise from non-equivalent system states. A fourth level is proposed as a purely conceptual extension describing potential relationships between system states. This level is not implemented and is not derived from experimental data. These observations are strictly exploratory and non-causal. The proposed framework does not establish mechanistic, predictive, or directional relationships, but provides a structured approach for analyzing constraint-driven systems across multiple levels of representation.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Desk Editor's Note private letter to a colleague

Single-case gait study finds similar performance but different state-space and embedding patterns under bite constraint, yet the numbers and checks are missing so the dissociation stays suggestive.

read the letter

The main point is that in this one Parkinson's patient, conditions with matched observable gait scores still produced visibly different trajectories in state space and different clusters in the unsupervised embedding when vertical occlusion was altered. That dissociation between aggregated performance and multi-level organization is what the authors put forward as the result worth noticing. They structure the analysis into linear metrics, dynamical-systems state space, and latent embeddings on the same raw data, and they keep the framing explicitly exploratory and non-causal, which is the right tone for the design. The multi-level lens itself is a reasonable way to ask whether simple output scores are enough to describe how the system reorganizes under a mechanical constraint. That part is clear and internally consistent. The soft spots are straightforward and fairly large. It is a single subject with repeated conditions but no reported trial counts, no statistical tests, no error bars, and no sensitivity checks on delay-embedding parameters, embedding method, or variable choice. The differences in the trajectories and latent space could therefore reflect session-to-session variability or analysis artifacts rather than genuine shifts in neuromechanical organization. The abstract supplies none of the quantitative detail needed to judge robustness, so the central claim rests on visual or qualitative contrast alone. This paper is for people already working on gait dynamics in constrained or clinical settings who are open to exploratory multi-representation approaches. A reader looking for solid evidence that performance metrics are insufficient would come away wanting more data and more controls. I would send it to peer review only if the authors add at least basic quantification, replication across sessions or subjects, and explicit checks on the embedding choices; without those the work stays too preliminary for a strong claim.

Referee Report

2 major / 1 minor

Summary. The manuscript claims that observable performance metrics do not fully capture underlying system organization in adaptive neuromechanical systems. Using a single-case design in a Parkinson's patient under repeated vertical dimension of occlusion (VDO) constraints, it compares three levels of analysis—aggregated linear performance metrics, state-space dynamical trajectories, and unsupervised latent embeddings—finding that conditions with comparable observable gait performance can exhibit distinct organizations in state space and latent space. The work is explicitly framed as exploratory and non-causal, with a fourth conceptual level proposed but not implemented or derived from data.

Significance. If the dissociation is substantiated, the result would underscore limitations of relying solely on aggregated metrics for inferring system states in biomechanical and neurological contexts, supporting multi-representation frameworks that combine dynamical systems and unsupervised embeddings. The transparency regarding the exploratory, non-causal scope and the single-subject intra-individual focus are strengths that align with the modest claims made.

major comments (2)

[Abstract] Abstract: The abstract presents an observational dissociation but supplies no quantitative details on trial counts, statistical tests, embedding validation, error bars, or sensitivity to analysis choices; this is load-bearing for the central claim that comparable observable performance corresponds to different organizations in state space and latent space, as the dissociation cannot be evaluated without these elements.
[Abstract] Abstract and implied methods: The single-subject exploratory design leaves differences in state-space trajectories and unsupervised latent embeddings vulnerable to session variability, non-stationarity, or embedding hyperparameter choices (e.g., delay, dimension, t-SNE/UMAP/autoencoder settings), without reported robustness checks or sensitivity analyses; this directly affects whether the observed dissociation reflects genuine neuromechanical organization rather than analysis artifacts.

minor comments (1)

[Abstract] Abstract: The statement that the fourth level 'is not implemented and is not derived from experimental data' should be moved or emphasized earlier to prevent any misreading of the empirical scope.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for their constructive comments on our exploratory single-case manuscript. We address the two major comments point by point below, agreeing where the concerns are valid and making targeted revisions to improve transparency without altering the non-causal, intra-individual scope of the work.

read point-by-point responses

Referee: [Abstract] Abstract: The abstract presents an observational dissociation but supplies no quantitative details on trial counts, statistical tests, embedding validation, error bars, or sensitivity to analysis choices; this is load-bearing for the central claim that comparable observable performance corresponds to different organizations in state space and latent space, as the dissociation cannot be evaluated without these elements.

Authors: We agree that the abstract should supply more concrete details to allow evaluation of the observed dissociation. As the study is explicitly exploratory and non-causal with no pre-specified hypothesis tests, no statistical tests, p-values, or error bars are appropriate or reported. We have revised the abstract to state the number of repeated gait trials per VDO condition and to specify the embedding parameters (delay, dimension) and unsupervised method settings used. The dissociation remains framed as an observational pattern in the structure of trajectories and embeddings rather than a statistically tested effect. revision: yes
Referee: [Abstract] Abstract and implied methods: The single-subject exploratory design leaves differences in state-space trajectories and unsupervised latent embeddings vulnerable to session variability, non-stationarity, or embedding hyperparameter choices (e.g., delay, dimension, t-SNE/UMAP/autoencoder settings), without reported robustness checks or sensitivity analyses; this directly affects whether the observed dissociation reflects genuine neuromechanical organization rather than analysis artifacts.

Authors: We accept that single-subject designs are susceptible to session effects and hyperparameter sensitivity. We have expanded the methods section to document the exact reconstruction and embedding parameters chosen and added a supplementary sensitivity analysis demonstrating that the reported dissociation in latent-space organization is stable under modest variations of delay, dimension, and perplexity. We have also strengthened the discussion to explicitly note non-stationarity and session variability as inherent limitations of the intra-individual design, reinforcing that the findings are observational and not claimed to be robust to generalization or causal inference. revision: partial

Circularity Check

0 steps flagged

No significant circularity; independent multi-level representations computed directly from raw data.

full rationale

The paper's central claim rests on comparing three independently computed representations—aggregated linear performance metrics, state-space trajectories via dynamical systems methods, and unsupervised latent embeddings—applied to the same raw gait data under repeated occlusal conditions. No equations reduce any result to a fitted parameter by construction, no self-citations serve as load-bearing justifications for uniqueness or ansatzes, and the fourth conceptual level is explicitly stated as unimplemented and non-derived from data. The analysis is labeled exploratory and non-causal, with the dissociation presented as an empirical observation rather than a derived prediction. This satisfies the criteria for a self-contained derivation without circular reduction.

Axiom & Free-Parameter Ledger

0 free parameters · 2 axioms · 0 invented entities

The analysis relies on standard dynamical-systems and unsupervised-embedding techniques whose interpretive validity is assumed rather than derived; the conceptual fourth level is explicitly noted as non-empirical.

axioms (2)

domain assumption Unsupervised embeddings of gait time series capture meaningful differences in system organization
The paper interprets latent-space differences as evidence of non-equivalent states without independent validation of the embedding.
domain assumption State-space trajectories constructed from gait variables reflect the temporal organization of the neuromechanical system
Standard assumption in dynamical systems analysis of movement; the paper does not test alternative variable sets.

pith-pipeline@v0.9.0 · 5542 in / 1495 out tokens · 40998 ms · 2026-05-09T20:11:46.524490+00:00 · methodology

discussion (0)

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

From Organization to Viability: A Multi-Level Analysis of Gait Dynamics Under Occlusal Constraint
q-bio.OT 2026-05 unverdicted novelty 5.0

In one Parkinson patient, higher occlusion produced the smallest longitudinal shift in PCA gait latent space over 11 weeks while immediate performance stayed comparable, supporting a viability level focused on sustain...

Reference graph

Works this paper leans on

29 extracted references · 21 canonical work pages · cited by 1 Pith paper · 1 internal anchor

[1]

Dental occlusion and posture: an overview

Michelotti A, Buonocore G, Manzo P, Pellegrino G, Farella M. Dental occlusion and posture: an overview. Prog Orthod. 2011;12(1):53–58. doi:10.1016/j.pio.2010.09.010

work page doi:10.1016/j.pio.2010.09.010 2011
[2]

Evaluation of the correlation between dental occlusion and posture

Baldini A, Nota A, Tripodi D, Longoni S, Cozza P. Evaluation of the correlation between dental occlusion and posture. Clinics (Sao Paulo). 2013;68(1):45–49. doi:10.6061/clinics/2013(01)OA07

work page doi:10.6061/clinics/2013(01)oa07 2013
[3]

Dental occlusion and body posture: no detectable correlation

Perinetti G. Dental occlusion and body posture: no detectable correlation. Gait Posture. 2006;24(2):165–168. doi:10.1016/j.gaitpost.2005.08.004

work page doi:10.1016/j.gaitpost.2005.08.004 2006
[4]

Dynamic Patterns: The Self-Organization of Brain and Behavior

Kelso JAS. Dynamic Patterns: The Self-Organization of Brain and Behavior. Cambridge (MA): MIT Press; 1995

1995
[5]

Dynamical Systems in Neuroscience

Izhikevich EM. Dynamical Systems in Neuroscience. Cambridge (MA): MIT Press; 2007

2007
[6]

Sensorimotor integration in human postural control

Peterka RJ. Sensorimotor integration in human postural control. J Neurophysiol. 2002;88(3):1097–

2002
[7]

doi:10.1152/jn.2002.88.3.1097

work page doi:10.1152/jn.2002.88.3.1097 2002
[8]

Postural control system

Massion J. Postural control system. Curr Opin Neurobiol. 1994;4(6):877–887. doi:10.1016/0959- 4388(94)90137-6

work page doi:10.1016/0959- 1994
[9]

A short latency vestibulomasseteric reflex evoked by electrical vestibular stimulation in healthy humans

Deriu F, Tolu E, Rothwell JC. A short latency vestibulomasseteric reflex evoked by electrical vestibular stimulation in healthy humans. J Physiol. 2003;553(Pt 1):267–279. doi:10.1113/jphysiol.2003.047274

work page doi:10.1113/jphysiol.2003.047274 2003
[10]

Unilateral trigeminal anaesthesia modifies postural control in human subjects

Gangloff P, Perrin PP. Unilateral trigeminal anaesthesia modifies postural control in human subjects. Neurosci Lett. 2002;330(2):179–182. doi:10.1016/S0304-3940(02)00779-6

work page doi:10.1016/s0304-3940(02)00779-6 2002
[11]

The relationship between the stomatognathic system and body posture

Cuccia AM, Caradonna C. The relationship between the stomatognathic system and body posture. Clinics (Sao Paulo). 2009;64(1):61–66. doi:10.1590/S1807-59322009000100011. 12

work page doi:10.1590/s1807-59322009000100011 2009
[12]

Is there a correlation between dental occlusion, postural stability and selected gait parameters in adults? Int J Environ Res Public Health

Nowak M, Golec J, Wieczorek A, Golec P. Is there a correlation between dental occlusion, postural stability and selected gait parameters in adults? Int J Environ Res Public Health. 2023;20(2):1652. doi:10.3390/ijerph20021652

work page doi:10.3390/ijerph20021652 2023
[13]

Exam- ination of the relationship between mandibular position and body posture

Sakaguchi K, Mehta NR, Abdallah EF, Forgione AG, Hirayama H, Kawasaki T, Yokoyama A. Exam- ination of the relationship between mandibular position and body posture. CRANIO. 2007;25(4):237–

2007
[14]

doi:10.1179/crn.2007.037

work page doi:10.1179/crn.2007.037 2007
[15]

Human balance and posture control during standing and walking

Winter DA. Human balance and posture control during standing and walking. Gait Posture. 1995;3(4):193–214. doi:10.1016/0966-6362(96)82849-9

work page doi:10.1016/0966-6362(96)82849-9 1995
[16]

Postural orientation and equilibrium: what do we need to know about neural control of balance to prevent falls? Age Ageing

Horak FB. Postural orientation and equilibrium. Age Ageing. 2006;35(Suppl 2):ii7–ii11. doi:10.1093/ageing/afl077

work page doi:10.1093/ageing/afl077 2006
[17]

Human postural control

Ivanenko YP, Gurfinkel VS. Human postural control. Front Neurosci. 2018;12:171. doi:10.3389/fnins.2018.00171

work page doi:10.3389/fnins.2018.00171 2018
[18]

Representation learning: a review and new perspectives

Bengio Y , Courville A, Vincent P. Representation learning: a review and new perspectives. IEEE Trans Pattern Anal Mach Intell. 2013;35(8):1798–1828. doi:10.1109/TPAMI.2013.50

work page doi:10.1109/tpami.2013.50 2013
[19]

Photorealistic Text-to-Image Diffusion Models with Deep Language Understanding

LeCun Y . A path towards autonomous machine intelligence. arXiv. 2022. doi:10.48550/arXiv.2205.11487

work page internal anchor Pith review doi:10.48550/arxiv.2205.11487 2022
[20]

UMAP: Uniform manifold approximation and projection for dimension reduction.Journal of Open Source Software, 3(29):861, 2018

McInnes L, Healy J, Melville J. UMAP: Uniform Manifold Approximation and Projection. J Open Source Softw. 2018;3(29):861. doi:10.21105/joss.00861

work page doi:10.21105/joss.00861 2018
[21]

Single-case experimental designs: a systematic review

Smith JD. Single-case experimental designs: a systematic review. Psychol Methods. 2012;17(4):510–

2012
[22]

doi:10.1037/a0029312

work page doi:10.1037/a0029312
[23]

Single-case research designs: methods for clinical and applied settings

Kazdin AE. Single-case research designs: methods for clinical and applied settings. 2nd ed. New York: Oxford University Press; 2011

2011
[24]

Effectiveness of a structured sophrology program on mental health symptoms

van Rangelrooij F, Solans-Buxeda A, Fernandez-Garcia M, Caycedo C, Selvam R, Bulbena A. Effectiveness of a structured sophrology program on mental health symptoms. Actas Esp Psiquiatr. 2020;48(1):16–27

2020
[25]

Vertical dimension: a study of clinical rest position and jaw muscle activity

Rugh JD, Drago CJ. Vertical dimension: a study of clinical rest position and jaw muscle activity. J Prosthet Dent. 1981;45(6):670–675. doi:10.1016/0022-3913(81)90426-1

work page doi:10.1016/0022-3913(81)90426-1 1981
[26]

Functional Occlusion: From TMJ to Smile Design

Dawson PE. Functional Occlusion: From TMJ to Smile Design. St Louis: Mosby; 2007

2007
[27]

Free-living gait characteristics in ageing and Parkinson’s disease: impact of environment and ambulatory bout length

Del Din S, Godfrey A, Galna B, Lord S, Rochester L. Free-living gait characteristics in ageing and Parkinson’s disease: impact of environment and ambulatory bout length. J Neuroeng Rehabil. 2016;13:46. doi:10.1186/s12984-016-0154-5

work page doi:10.1186/s12984-016-0154-5 2016
[28]

Performance of instrumented insoles for gait analysis

Martin E, Leboeuf F, Pradon D. Performance of instrumented insoles for gait analysis. Sensors (Basel). 2024;24(18):6043. doi:10.3390/s24186043

work page doi:10.3390/s24186043 2024
[29]

Test-retest reliability of gait parameters measured with instrumented insoles in people with Parkinson’s disease

Mostovoy A, Jacobs D, Farid L, Dhellin P, Baille G. Test-retest reliability of gait parameters using instrumented insoles. PLOS Digit Health. 2023;2(11):e0000262. doi:10.1371/journal.pdig.0000262. 13

work page doi:10.1371/journal.pdig.0000262 2023