8.8 KiB
Academic Wording Guidelines for StarForth
Version: 1.0 Date: 2025-12-14 Purpose: Maintain academic rigor and avoid causation claims
Critical Rule: Avoid Implied Causation
When noting mathematical similarities between StarForth's empirical results and equations from physics, always use language that describes structural similarity, not equivalence.
✅ APPROVED PHRASES (Use These)
Describing Mathematical Relationships
Best (Most Defensive):
"We evaluated multiple candidate models and found that the functional form
τ = τ₀/√(1-β²) provided the best empirical fit across workloads."
Good:
- "fits the functional form"
- "shares structural similarity with"
- "is mathematically isomorphic to"
- "exhibits the same mathematical structure as"
- "the empirical data conforms to"
- "matches the pattern of"
Describing Observations
Good:
- "We observed that..."
- "The data reveals..."
- "Measurements indicate..."
- "Empirical analysis shows..."
- "The experimental results demonstrate..."
Noting Physics Parallels
Good:
- "shares dimensional similarity with"
- "the functional form also appears in [physics context]"
- "this mathematical structure is also found in"
- "bears mathematical resemblance to"
❌ FORBIDDEN PHRASES (Never Use)
Direct Equivalence Claims
Never:
- "is equivalent to"
- "is the same as"
- "proves that"
- "demonstrates that [physics concept] applies"
Unqualified Analogies
Avoid (unless in clearly-marked interpretation sections):
- "is analogous to"
- "corresponds to" (without "structurally" or similar qualifier)
- "behaves as" (without qualification)
- "mirrors" (implies direct correspondence)
- "exactly as predicted by"
Causation Implications
Never:
- "because of [physics principle]"
- "due to [physics concept]"
- "governed by [physics law]"
- "obeys [physics equation]"
Examples: Before & After
Example 1: Performance Scaling
❌ WRONG:
"Performance scales exactly as predicted by special relativity's time dilation."
✅ CORRECT:
"We evaluated multiple candidate performance models. The empirical data fits
the functional form τ = τ₀/√(1-β²), which shares structural similarity with
the Lorentz transformation from special relativity."
Why: The correct version:
- States we tried multiple models (not cherry-picking)
- Uses "fits the functional form" (empirical observation)
- Uses "shares structural similarity" (mathematical fact, not causation)
Example 2: Window Capacity Law
❌ WRONG:
"The system obeys a cosmological constant law Λ(DoF) = 4096/(DoF+1)."
✅ CORRECT:
"Empirical measurements reveal that window capacity follows the relationship
Λ(DoF) = 4096/(DoF+1) with 0.00% coefficient of variation. The symbol Λ is
borrowed from cosmology for mathematical convenience."
Why: The correct version:
- "Empirical measurements reveal" (observation, not theory)
- "follows the relationship" (describes pattern, not obedience to law)
- Explicitly states Λ is "borrowed... for mathematical convenience" (notation, not physics)
Example 3: Geodesic Convergence
❌ WRONG:
"Workloads converge along geodesics analogous to general relativity."
✅ CORRECT:
"All workload trajectories converge to the same attractor basin regardless of
starting conditions. This pattern shares structural similarity with geodesic
motion in curved spacetime."
Why: The correct version:
- Describes observed behavior first (empirical fact)
- "shares structural similarity" (mathematical comparison, not physical claim)
- No causation implied
Defensive Language Patterns
Pattern 1: Model Evaluation Statement
Template:
"We evaluated [N] candidate models including [list]. The functional form
[equation] provided the best fit with R² = [value]."
Example:
"We evaluated linear, polynomial, exponential, and power-law models. The
functional form f(t) = f₀×e^(-λt) provided the best fit with R² = 0.96."
Defense: Shows you didn't cherry-pick; you tested alternatives.
Pattern 2: Empirical Observation + Mathematical Note
Template:
"[Empirical observation]. This [relationship/pattern/structure] also appears
in [context] as [reference]."
Example:
"Performance scales nonlinearly with degrees of freedom. This functional form
also appears in special relativity as the Lorentz factor."
Defense: Observation is separate from physics reference; no causation implied.
Pattern 3: Metaphor Disclosure
Template:
"We employ [physics concept] as a conceptual metaphor, where [mapping]. The
actual implementation uses [literal description]."
Example:
"We employ thermodynamic cooling as a conceptual metaphor, where execution
frequency decreases over time. The actual implementation uses exponential
decay f(t) = f₀×e^(-λt) applied periodically."
Defense: Explicitly labels metaphor as metaphor; provides literal implementation.
Section-Specific Guidelines
In Abstracts/Titles
Use: Mathematical/empirical language only
- ✅ "Frequency-Based Adaptive Runtime"
- ❌ "Physics-Driven Adaptive Runtime"
In Introductions
First: State empirical observations Then: Note mathematical similarities Never: Claim physics explains computation
In Methods/Implementation
Use: Literal descriptions only
- ✅ "Frequency counter incremented on execution"
- ❌ "Temperature increases when word heats up"
In Results
Use: Statistical language
- ✅ "CV = 0.00% (p < 10^-30)"
- ❌ "Perfect thermodynamic equilibrium achieved"
In Discussion/Interpretation Sections
Allowed: Exploratory language, IF properly qualified
- ✅ "One interpretation is that... However, causation is not established."
- ✅ "This pattern suggests possible structural similarities with..."
- ❌ "This proves that computation follows physics laws"
Academic Reviewer Defense Scenarios
Scenario 1: "You're claiming physics!"
Response: "No. Section [X] explicitly states: 'We employ [concept] as a conceptual metaphor.' All empirical claims are mathematical: the functional form f(t) = f₀×e^(-λt) fits our performance data with R² = [value]. That this same functional form appears in physics is noted as a mathematical similarity, not a physical claim."
Scenario 2: "You cherry-picked equations!"
Response: "Section [X] documents that we evaluated [N] candidate models: [list]. The reported functional form was selected based on empirical fit quality (R² = [value], AIC = [value]), not theoretical preference."
Scenario 3: "This is pseudoscience!"
Response: "All claims are empirically verifiable:
- [Claim 1]: See [data file], line [X]
- [Claim 2]: See [experimental results]
- [Claim 3]: Reproduction protocol in [file]
The physics references describe mathematical similarities, clearly documented in ONTOLOGY.md Section [X] as metaphorical framing, not physical causation."
Checklist Before Publication
Before submitting any paper, presentation, or public documentation:
- Search for "is equivalent to" → replace with "fits the functional form"
- Search for "is analogous to" → replace with "shares structural similarity"
- Search for "corresponds to" → add qualifier like "structurally"
- Search for "mirrors" → replace with "matches the pattern of"
- Search for "exactly as" → replace with "according to the functional form"
- Search for "obeys" → replace with "follows the relationship"
- Search for "governed by" → replace with "characterized by"
- Verify all metaphors are explicitly labeled in text
- Confirm empirical claims have data references
- Check that all physics parallels use "shares similarity" language
Cross-Reference
See also:
- ONTOLOGY.md - Section 3.2 "Avoid/Deprecated Terms"
- ONTOLOGY.md - Section 7.1 "Academic Writing Guidelines"
- ONTOLOGY.md - Section 1.2 "Conceptual Relationships" (metaphor mapping)
Summary
Golden Rule: Describe what you observed, note where it mathematically resembles known equations, but never claim the physics causes the behavior.
Your Position:
- "We measured these relationships in our system"
- "They fit these mathematical forms"
- "These forms also appear in physics"
- "This similarity may provide useful modeling frameworks"
NOT Your Position:
- "Our system implements physics"
- "Physics laws govern our system"
- "This proves computation is physical"
Maintain this discipline rigorously, and no academic reviewer can challenge your scientific integrity.
License: See ../LICENSE