The three proposal texts were reviewed paragraph-by-paragraph and rewritten in a direct, mechanics-focused research voice. Long reviewer-facing constructions were replaced with the explicit sequence: evolutionary question → testable hypothesis → experimental design → alternatives/nulls → interpretation. The tone aims to be clear and suitable for an introductory exchange with an evolutionary biologist. It does not claim to reproduce a private author’s distinctive style exactly.
What changed
Missing muscles: Clear separation between published anatomical reconstruction and the proposed contribution: honest identification of recoverable soft-tissue properties; extant masked tests; phylogenetically held-out calibration.
Joint evolution: Explicit distinction between passive motion, load-achievable motion, force-generating potential and observed behavior; the seven-family hyperelastic atlas is a baseline, not a validated animal-tissue library.
Spinal function: Falsifiable order-invariance null under constant moment and independent serial joints; dependence on measured intact-column loading for physical validation.
There were 22 substantive paragraph/aim edits in each proposal (66 total). Reference lists and already-accurate headings were retained. The HTML beneath each corresponding demo was regenerated from the edited Word file and checked for paragraph-level equivalence.
Publication checks and accurate scope
Bishop & Pierce (2024), Anatomical Record forelimb and hindlimb series: DOIs 10.1002/ar.25312, 10.1002/ar.25310. The online-first dates are in 2023; the cited issue is 2024.
Bishop & Pierce (2026), Bulletin of the Museum of Comparative Zoology, 164(2), 145–184: DOI 10.3099/MCZ85.1.
Demuth et al. (2022), Scientific Reports: DOI 10.1038/s41598-022-07074-x. Published 3D muscle reconstructions were checked, but they do not constitute a balanced multi-clade atlas for this proposal.
Bishop & Pierce (2024), Science Advances: DOI 10.1126/sciadv.adr2722. Their inference of late erect hindlimb function in synapsids is properly acknowledged rather than claimed as new.
Wright & Pierce (2026), Journal of Experimental Biology: DOI 10.1242/jeb.251911. The tegu/opossum hip tissue-removal and XROMM study supplies a basis for an experimental comparison; specimen-level access is not assumed.
Jones, Angielczyk & Pierce (2024), Proceedings of the Royal Society B: DOI 10.1098/rspb.2024.0820; associated Dryad 10.5061/dryad.2547d7wx9. The archive includes digital-bending/simulation and morphology data, not independently measured intact-column loading.
Asadi & Laksari (2026), PI-UNet article: DOI 10.1007/s10439-026-04321-4; synthetic 3D benchmarks, not validation on fossils.
Scientific limits still applicable
None of the interactive demos uses measured fossil muscle tissue or experimentally validated extinct locomotion. They are pedagogical surrogates.
A conditional reconstruction set depends on sampled extant taxa, anatomical homology and fossil preservation. Interval coverage is an empirical test, not a given.
Bone observations do not uniquely identify tissue law, activation, pennation or maximum muscle force.
A passive pose envelope should not be interpreted as an animal’s actual gait.
A mechanical change is not, by itself, evidence of historical natural selection or adaptation.
No lab datasets, specimens, collaboration agreements, or external experimental validation are secured by these files.
Technical verification
All three proposals render as two-page DOCX files without clipping.
All proposal paragraphs are present verbatim in the corresponding online-ready HTML section.
Chromium checks cover desktop and mobile with zero horizontal overflow and no JavaScript errors.
Joint counterfactual output changes under material alteration.
Spine model returns identical total angle at its zero-coupling null.
Landing page includes the supplied pixel-sunglasses portrait, verified research links, and social metadata.
Deployment
Upload the extracted research-directions/ folder to the root of the GitHub Pages repository, replacing the prior folder. This package was not pushed to GitHub; verify the updated pages after deployment.