H-29 - A metastable negatively charged strangelet stopped in matter would grow exponentially by alternating nuclear absorption and electron capture
The danger hypothesis whose probability the whole strangelet analysis prices. The paper’s own candidate terminators of the growth, none of which QCD knowledge can confirm or exclude: (a) a stable positively-charged species may be reached during growth, which electron shielding then freezes; (b) the lightest metastable strangelet may be isolated by many baryon-number units from the next stable one (as the A=8 gap stops alpha-alpha fusion in stars), so growth by single-nucleon absorption cannot proceed; (c) capture-energy release may fragment the strangelet into small unstable pieces. Growth would continue at least until released energy vaporizes surrounding material away. Note the scenario requires production velocity below ~0.1c (faster strangelets break up before stopping), hence lifetime > ~1e-7 s to traverse detector and shielding.
Link to original
H-40 - A stopped negative strangelet would not grow without limit
Constructed complement minted at step 4b: the corpus describes the runaway-growth mechanism but no paper states the self-limiting alternative, which is the logically needed competing answer.
Link to original
The runaway mechanism against its self-limiting complement, both conditional on the strangelet existing, stopping, and being negatively charged in bulk. Exhaustive by construction: proposition and logical complement. A failure of the analysis — e.g. the growth argument and its critiques sharing a wrong picture of the strangelet surface — would favour the second member, since the runaway claim is the one carrying the extraordinary conclusion.
Prior
# HC-8 prior — runaway growth of a stopped negative strangelet. Members in HC-8.hypotheses order:
# [H-29 runs away without limit, H-40 some barrier halts it]. exhaustive_by_construction: true, so there is
# no residual member (rule 4 does not apply): the two are a proposition and its exact logical negation.
#
# Conditioning (depends_on HC-7, prose not imported): every weight here is CONDITIONAL on the antecedent the
# cluster fixes — a (meta)stable strangelet was produced, stopped (v < 0.1c, lifetime > ~1e-7 s), and is
# NEGATIVELY charged in bulk. Whether that antecedent holds is HC-7's business (bulk stability) plus the
# production/charge conditions priced elsewhere; A-57's headline claim that stable strange matter is almost
# surely POSITIVE bears on that antecedent, so it is NOT re-used here to avoid double-counting the conditioning.
# Only the growth-vs-no-growth content of A-56/A-57 (approved, no-observation args reaching H-29 via
# affects_hypotheses; rule 6) enters as reasoning below.
#
# Evidence split (rule 5): HC-8's three inbound E edges — E-30 (O-23 lunar persistence), E-31 (O-39 cosmic-ray
# collision counts), E-32 (O-22 stars/Sun/WD/NS persist) — are all DISCRIMINATING survival bounds (each edge's
# "why this is evidence" points toward H-40 with strength scaling with the per-collision production-and-stopping
# probability). None is an outside-view base rate the prior should rest on, so NONE is marked used_for_prior;
# all three are left for step 8. This prior therefore marks no edges.
# --- H-29 (runaway): the danger hypothesis. What would have to be true: given the worst-case antecedent, the
# absorb-nucleus / electron-capture cycle meets NO barrier across ~50 orders of magnitude of baryon-number
# growth. BJSW's own text supports only the weak "we know of no absolute barrier" — an absence-of-knowledge
# claim, not a positive case — while listing three concrete candidate terminators. Priced well below H-40. ---
# Outside view first (rule 2). Reference class: proposed self-amplifying / unbounded-runaway catastrophe
# mechanisms in physics whose proposers could not name a barrier. Degenerate/bound many-body systems generically
# grow a surface, Coulomb, or charge barrier as they scale, so "truly no barrier anywhere" is the rare outcome.
p_no_barrier_outside = 0.10 # base rate that a proposed unbounded runaway has no self-limiting barrier at all
# Mechanistic view: the three BJSW terminators (H-29 note a/b/c; A-56 wrinkle 3a/3b; A-57), each a chance to
# halt growth. Probabilities below are that the terminator does NOT fire (growth survives it). A-56 flags these
# "cannot be modeled accurately enough to carry the safety case", so each is plausible-but-uncertain, not ~1.
p_no_chargefreeze = 0.45 # (a) A-57: Fermi-gas kinematics drives Z/A positive as A climbs; a stable positive
# species then freezes growth via electron shielding. Likely to fire -> 0.45 survives.
p_no_gap = 0.60 # (b) A-56 3b: a stability gap (like the A=8 gap halting alpha-alpha fusion) can isolate
# the metastable strangelet so single-nucleon climb stalls; plausible, not assured.
p_no_fragment = 0.70 # (c) capture-energy release may shatter the strangelet into small unstable pieces;
# the least compelling of the three, so most mass survives it.
p_runaway_mech = p_no_chargefreeze * p_no_gap * p_no_fragment # all three terminators fail -> no barrier
# Two frameworks -> inline mixture (rule 7). c_mech is the credence the three named terminators exhaust the
# barrier space; the outside view warns there are likely MORE barriers than the three modelled, pulling lower.
c_mech = 0.60 # credence the explicit-terminator decomposition captures the full set of halting mechanisms
p_runaway = c_mech * p_runaway_mech + (1 - c_mech) * p_no_barrier_outside
# --- H-40 (no runaway): some barrier halts growth before macroscopic conversion. The outside-view-favoured and
# BJSW-flagged outcome; carries the bulk of the conditional mass. Anchored at unit weight; H-29 is priced as
# odds against it, so a skeptic overrides p_runaway (or any terminator factor) and both weights move together. ---
w_no_runaway = 1.0
w_runaway = p_runaway / (1 - p_runaway) # odds of runaway vs. not; arithmetic identity over H-29's own p
prior("HC-8", [w_runaway, w_no_runaway])