14 October 2026 · Complete Day 2 analysis
Clearer accounts.
Harder questions.
Six Cazorla Jr passengers add detail about the journey, a life jacket and the struggle to reach safety. Their testimony also exposes a critical distinction: what someone saw, what they assumed, and what records must still establish.
On this pageWhat changed
01 / The update
More evidence about safety and survival. Still no complete collision reconstruction.
The strongest additions concern Cazorla’s occupants, preparation and aftermath. They do not supply a new eyewitness view of Ocean Angel’s approach.
A jacket was worn.
Williams describes Taylor helping him put it on before the journey resumed. That improves the Day 1 position, which established a request and retrieval but left wearing less clear.
Tyson helped the couple.
Taylor and Williams describe his assistance from the water to floating wreckage. Edward adds a supporting account. This supports the rescue action without deciding collision responsibility.
The slowdown gains support.
Williams says Pascal’s instruction was followed immediately and the reduction maintained. Chris also says Tyson complied immediately. Neither measures the resulting speed.
Bow seat ≠ assigned lookout.
Butcher denies the lookout role. Chris eventually identifies the seat itself as his reason for assuming the role. Lester does not know whether an appointment was made.
Training ≠ operating licence.
Edward’s examination exposes confusion between a completed course and authority to operate. The chair orders the documents, with a specific place and deadline for delivery.
The sinking gets more detail.
Lester describes impaired control, jacket retrieval and flooding. His hatch account is challenged against the recovered wreck. The image and salvage record must test that difference.
What remains unchanged: none of these six witnesses supplies a direct account of watching the other boat approach. Their failure to see its lights does not establish that Ocean Angel was unlit. Day 2 adds no authenticated registry resolution, measured impact speed or defensible collision angle. The vessel mockups and scenario visuals remain illustrations, not recovered evidence.
02 / People and knowledge
Six witnesses. Six different windows onto the same journey.
Presence, spelling and operational role need separate confidence judgments. Expand each account to see what it adds and where it stops.
Working names follow the broader record. Riviere/Rivier and Romeon/Romaine vary between the transcript and reporting; Williams’s final “s” is expressly clarified. Butcher’s first name and Chris’s surname remain unresolved. Lester/Soya and Butcher/Mavado are different people. Explore the full name and boat register →
01Part I · A participant in the rescueTaylor RiviereDescribes the jacket request for her companion, being thrown into the water, an injured leg and Tyson’s help reaching floating wreckage.
Taylor identifies illuminated red and green lights at Cazorla’s bow. She describes Tyson swimming to her and Williams and helping them reach a piece of Ocean Angel. After slipping off it, she remained alongside. Her account continues through a rescue boat, a landing at Canaries and treatment after leaving the beach. She expressly distinguishes the people taken on that rescue trip from Dylan and Tyson, who remained at the wreckage.
Limit of the account. She did not see Ocean Angel approach and says she cannot know whether its lights were on. Her identification of an unresponsive woman as Miss Nestor needs reconciliation with other rescue-load accounts. Her injury does not establish a particular impact mechanism.
Link to this witness02Part I · The passenger wearing the requested jacketRomeon / Romaine WilliamsConfirms putting on the jacket with Taylor’s help, the immediate slowdown after Pascal’s instruction, and Tyson’s assistance in the water.
Williams says he was a non-swimmer on his first boat trip. He describes Cazorla stopping for the jacket, resuming, and the impact a few minutes later. He says the reduced speed was maintained. After reaching wreckage he called Taylor’s father and describes phone-light signaling, the rescuer’s slow approach, Canaries and pickup transport to Soufrière. He says he was uninjured and accompanied Taylor onward to St Jude.
Limit of the account. He did not watch the other craft approach, cannot give a measured speed and cannot tell whether the engine stopped during the jacket pause. His reasons for requesting the jacket should remain his own account; repeatedly suggesting fear caused by speed does not establish that explanation.
03Part II · Presence established; credentials require documentsJean Pierre Ralph EdwardPlaces himself aboard the return trip, describes Tyson helping Taylor and her partner, and recounts a second immersion and evacuation through Ti Kaye.
Edward says he boarded as a passenger with a friend called Wabi or Wabby. He reports darkness, no rain and some chop, then an impact without a prior sighting. He says he reboarded, saw his friend injured, and later swam to a yacht after Cazorla flipped. His examination eventually distinguishes a course certificate from an operating licence, and the chair orders the relevant documents delivered by a fixed deadline.
Limit of the account. His seating side, wearer count, work history and certification answers are uneven. Red/green terminology is confused; that is not proof the actual lights were installed backwards. His suggested glare explanation is his inference, not something the bow passenger told him caused the collision.
04Part II · Bow passenger; lookout role disputedButcher, known as MavadoConfirms sitting forward but denies being the lookout. His account of who wore the life jacket changes from certainty to an acknowledged assumption.
Butcher says he was given a ride and sat at the bow. He recalls red and green lights aboard Cazorla and seeing no approaching craft. He first says a woman wore the jacket, then explains that he turned forward after passing it on and did not see who wore it. Questioners also confront him with his police statement about recognising Tyson at the helm and attempting to move the damaged boat shoreward.
Limit of the account. His denial does not by itself prove there was no lookout; other witnesses’ seat-based assumptions do not prove he accepted that duty. The transcript varies his first name. Confrontational questioning and inconsistent answers are not a toxicology result or proof of deliberate deception.
05Part II · Seating and the first seconds after impactChris, surname unresolvedPlaces himself beside Pascal and Lester, recalls a separate engine noise after impact, and explains the assumptions behind his jacket and lookout claims.
Chris says he sat left of Dwight Pascal, with Lester Jean to the right, ahead of the console. He says Tyson slowed immediately when asked, and red, green and white lights stayed on. He recalls a crash, then a rough engine sound, then Tyson stopping the engine. He heard help from his right. His claim that Butcher was lookout eventually rests on where Butcher sat; his belief that jackets were sufficient rests on trust in the owner, without a count.
Limit of the account. He saw nothing strike the engine. The second sound does not identify a propeller contact, a person or a particular damaged component. His surname appears as Jabatis and other variants in the automated text; no expansion is treated as verified.
06Part III · Flooding, jackets and the sinkingLester Jean, known as SoyaAdds the fullest new account of efforts aboard the damaged Cazorla: steering difficulty, jacket retrieval, sinking and transfer through Ti Kaye.
Lester describes a changed passenger complement on the return journey. After impact he says Pascal tried to reach shore while Mavado helped with an unstable engine arrangement. He says he opened a bow hatch and retrieved jackets as the boat flooded, then sought help at a yacht. He describes a later rescue boat, Ti Kaye, and another boat to Soufrière. He does not claim to have operated Cazorla on this trip.
Limit of the account. The recovered boat’s condition is put to him as a challenge to the hatch account. That dispute needs dated photographs and a recovery history. He did not observe the approaching craft or individual ejections; distance estimates and engine descriptions are too uncertain for a precise reconstruction.
03 / After the impact
Two rescue routes must remain distinguishable.
These are reported sequences, not a common clock or a complete accounting of everyone rescued. A person reaching another boat’s wreckage did not necessarily depart on that boat.
Taylor and Williams’s accounts
Water → wreckage → Canaries
- Immediate assistanceTyson swims to the couple and helps them reach a floating piece of Ocean Angel.
- Contact and signalingWilliams calls Taylor’s father. Phone lights help attract a rescue boat; witnesses describe its slow approach.
- A separate rescue loadThe couple go to Canaries. They say Dylan and Tyson remain at the wreckage for later retrieval.
- Medical transferTaylor and Williams leave Canaries by pickup for Soufrière; Williams accompanies Taylor onward to St Jude.
Edward and Lester’s accounts
Damaged boat → yacht → Ti Kaye
- A developing emergencyPeople remain or reboard as Cazorla takes water. Accounts differ over control, engine state and movement.
- Another entry into the waterThe boat sinks or flips in the descriptions. Survivors seek refuge or help at a nearby yacht.
- Transfer to Ti KayeA rescue boat brings members of this group to the shore facility. They describe waiting for assistance.
- Onward transportAnother boat takes survivors toward Soufrière. Dispatch and hospital records are needed to establish times.
The practical question is no longer only who rescued whom. It is how each person moved from water or wreckage to a rescue craft, a landing and medical care. That record can test delays, prevent double-counting and establish which communications reached which service. A survivor’s estimate of a long wait is a serious lead; it does not itself reveal dispatch time, availability or the reason for a delay.
Follow accounts without blending them.
The account explorer separates witnesses and links scenes to their sources. The evidence explorer lets you test the individual claims and qualifications.
04 / The closer reading
Where the testimony changes—and why it matters.
An inconsistent answer deserves examination. It does not automatically establish dishonesty. Read the question, answer and later qualification together.
Who wore the jacket? Read through to the qualification.
One assertion is later explained as an assumption
Taylor and Williams describe the jacket being fitted on Williams. Butcher initially insists a woman wore it, then explains he did not turn back to watch and had assumed the recipient. Edward’s oral account names only Taylor’s partner, while a passage read from his police statement describes about two other wearers as well. These are different issues: the requested jacket’s wearer has strong first-person support; the total number wearing jackets still needs careful reconciliation.
Did Cazorla have a lookout? The seat does not settle the role.
A genuine cross-witness dispute
Pascal’s Day 1 account identifies Mavado as lookout. Butcher denies the assignment. Chris repeatedly calls him lookout, but ultimately explains that he inferred the duty from the bow seat. Lester says he does not know whether anyone appointed him. A passenger facing forward, an assigned lookout and an effective lookout are three different propositions. The next questions are who instructed whom, whether that role was accepted, what could be seen and how warnings were to be communicated.
Were there enough jackets? Confidence is not an inventory.
A proposed correction has an uncertain foundation
Chris asks to remove the clause in his police statement saying he could not tell whether there were enough jackets for everyone. Later questioning establishes that he did not count them or know how many were aboard. He relies on his belief that the owner is responsible and acknowledges the assumption, while continuing to assert sufficiency. The public record should preserve the original uncertainty, requested amendment and subsequent explanation together.
Was the damaged boat driven, drifting or unable to steer?
Sequence and terminology remain unresolved
Chris describes an engine running after impact until Tyson stopped it. Butcher says he reached the controls but could not drive because the engine was off. Edward’s police statement, read aloud, attributes a shoreward move to Mango. Lester describes attempts to move with damaged control and an unstable engine, while accepting some movement before sinking. These accounts may concern different moments. They need a common sequence before being classified as mutually exclusive. A question proposing that someone valued the boat over people is not evidence of that motive.
Could Lester open a hatch that was absent from the recovered hull?
A physical-evidence dispute
An examiner challenges the hatch account using the recovered boat’s damage. Lester maintains the cover existed when he retrieved jackets and raises the possibility of later damage. A recovered wreck is not automatically in precisely its immediate post-impact condition. The same compartment must be identified on pre-incident, scene and salvage images before the account can be tested fairly. Neither the examiner’s certainty nor the witness’s explanation is a completed structural investigation.
Do different names mean different people—or different rescue trips?
Identity and journey reconciliation needed
Taylor identifies an unresponsive woman in her rescue account as Miss Nestor; Elias’s Day 1 account names Debbie. Before treating those as incompatible, establish which rescue craft, load and landing each witness describes. Likewise, Wabi, Nyal and Maddak must not be fitted into a nine-person roster by elimination. Chris and Edward now establish their own presence, but that does not automatically resolve every nickname. The departure roster and the later rescue loads answer different questions.
Was there a safety briefing? Preserve the qualified minority account.
Different observations of departure
Taylor, Williams and other passengers describe no briefing. Edward later considers whether one could have preceded his late arrival. Butcher says he heard the captain talking when he arrived and calls it a briefing, but cannot explain its content. This does not establish an adequate safety briefing; it should still remain visible beside the contrary accounts. The practical question is what each boarding passenger was told about jackets, emergency action and the voyage—not whether someone was merely speaking.
Some apparent contradictions are really corrections or different moments.
Do not manufacture a credibility problem
No jackets at departure can coexist with a jacket fitted later. No approaching boat lights observed can coexist with visible shore or yacht lights. Taylor corrects a questioner who attributes Windjammer to her; she had not supplied it. Lester corrects confusion between what the captain did before and after impact. Williams corrects the suggestion that his Canaries departure was by ambulance: he and Taylor left by pickup. Preserve the corrected answer rather than repeat the mistaken premise.
05 / Assessing the process
The strongest questioning tests the basis of knowledge.
Saint Lucia benefits when the hearing makes assumptions visible, insists on documents and protects a fair process. Those strengths also provide a standard for scrutinising the questioning itself.
What deserves credit
A clear documentary next step
Chair Mario Michel’s examination moves beyond Edward repeatedly saying he is certified. It distinguishes a course from a licence and ends with a directive: relevant documents to Soufrière police by noon on Friday, 16 October. Naming the records, recipient and deadline makes follow-up possible.
Observation separated from assumption
The chair asks why Chris identifies Butcher as lookout and elicits the seat-based inference. Other questioning tests the uncounted jacket inventory. These exchanges explain why a confident answer can still have a weak factual foundation.
Fairness during difficult evidence
The chair explains the inquiry’s prevention purpose and encourages candour about operating practice. In the final examination, he limits discussion of a separate pending matter to avoid prejudicing the witness, while keeping attention on the rescue chronology. That procedural protection is a concrete strength.
Where discipline remains necessary
A proposition in a question is not a finding
Questions suggest an unlit approaching vessel, reasons a lookout failed to see it, and a motive for moving the damaged boat. Some treat the lookout denial as settling the role. Those propositions must be tested against the answers and independent evidence, rather than repeated as the Commission’s conclusions.
Keep later qualifications attached
A short clip of Butcher insisting a girl wore the jacket would omit his later explanation that he had not watched the wearer. A clip of Chris asserting sufficient jackets would omit the absence of a count. Responsible coverage needs the full evidential development.
Use physical evidence to test recollection
The hatch dispute is an opportunity to show the relevant photographs, identify the exact compartment and account for recovery damage. Calling an account incorrect is not a substitute for demonstrating why. The same standard applies to engine condition, sightlines, timing and registration history.
The automated text does not consistently identify speakers. This analysis attributes specific interventions to the chair where the exchange identifies that role; it does not assign every helpful or problematic question to him. Criticism of a question is not a determination about the questioner’s motives.
06 / The next evidence
Ten priorities for a record that can support conclusions.
These are SLPA’s follow-up questions, not Commission findings or assertions that the requested records exist. Each connects an uncertainty to evidence capable of resolving it.
01Reconcile each boat’s people, roles and rescue loads.
Obtain a name-by-name account for each outward and return leg. Separate captain, owner, passenger, assigned crew and later rescuer. Resolve aliases directly with the people or authenticated records, not by arithmetic.
Records or evidence to seek: Witness list, signed statements, boarding and departure records; a separate rescue craft/load/landing matrix.
02Establish who was assigned to keep lookout.
Ask each relevant person about the instruction, acceptance, position, continuous attention and warning method. Then test the actual sightlines, including any interruption during jacket retrieval.
Records or evidence to seek: Captain, owner and bow passenger accounts; documented vessel layout, bow trim and night visibility assessment.
03Authenticate both boats’ lighting and physical damage.
Separate lights fitted, lights powered and lights visible from another craft. Map damage, engine attachment and the disputed hatch using a dated recovery chain. Non-observation alone cannot answer these questions.
Records or evidence to seek: Original photographs, inspections, electrical and mechanical examination, salvage records and any expert reconstruction with its assumptions disclosed.
04Build a reliable clock before calculating a speed.
Anchor departure, slowdown, jacket stop, impact and sinking to records where available. Keep witnesses’ rough estimates visible. A journey average does not determine the speed at impact.
Records or evidence to seek: Call and device records, event departure material, VHF/lighthouse records, rescue logs, surveyed distances and confidence ranges.
05Verify operating authority, not just a course certificate.
Check the documents in force on the incident date, the person actually operating and the role being performed. Follow up the chair’s order for Edward’s records by noon on 16 October; the deadline is a documentary checkpoint, not proof of compliance.
Records or evidence to seek: SLASPA course, practical assessment, licence, endorsement, renewal and expiry records; the applicable requirements for each role and craft.
06Count the jackets and test access, fit and use.
Availability, sufficiency, accessibility and wearing are separate safety questions. A jacket behind a hatch is not the same as one already fitted when a person enters the sea.
Records or evidence to seek: Inventory and recovered equipment, inspection records, storage layout, individual wearer accounts and the briefing/fitting procedure.
07Reconstruct the post-impact decisions without assigning motives.
Identify who stayed aboard, who was thrown, who jumped to assist, who reboarded and who handled the controls. Test whether powered movement, drift, steering failure and flooding describe different stages.
Records or evidence to seek: Person-by-person sequence, engine and steering findings, rescue positions and contemporaneous photographs or communications.
08Reconcile the complete emergency-response chain.
Find the first verified notification and trace dispatch, launch, rescue, landing and medical transfer. Reports of waiting at Ti Kaye must be assessed alongside the separate Canaries route.
Records or evidence to seek: Emergency calls, maritime and ambulance dispatch records, hotel and landing accounts, rescue operators and receiving-hospital times; protect private contact and medical details.
09Close the statement-handling gaps.
The morning witnesses’ statements were unavailable for their examination. Williams says a statement was typed and read back but not signed. Compare that with the earlier investigator’s account before drawing conclusions about how the records were handled.
Records or evidence to seek: Original statement files, recording and readback dates, signatures/adoption, custody history and the precise amendments recorded during this hearing.
10Test whether safety systems work for every passenger trip.
Passengers describe uneven briefings and limited jacket wearing on local leisure travel. Accounts of historic boating work raise further training and enforcement questions. These are leads to investigate, not a census of operators or proof of current employer wrongdoing.
Records or evidence to seek: Applicable duties by vessel/trip class, operator and regulator responses, briefing checks, inspection and enforcement records, practical training and consistent reporting arrangements.
07 / From evidence to reform
Better rules need an operating system behind them.
The inquiry’s remit includes the collision and the maritime laws, procedures and safety arrangements around it. A policy response must distinguish missing rules from rules that are not understood, checked or applied.
Day 2 makes three reform questions especially concrete: how a passenger knows that a named person is qualified to operate; how a voyage begins with a usable briefing and fitted safety equipment; and how a distress call becomes a coordinated rescue and medical transfer. These questions concern local trips as well as tourist excursions. Witness experience can expose a problem, but a wider conclusion requires operator and regulator records.
SLASPA’s connection is documentary and operational. The hearing repeatedly turns to its training and licensing processes, boat certification and the oversight witnesses say they encountered. Edward’s record request is one tangible test. Lester’s account of historic boating work raises further leads, with dates and duties requiring corroboration. This coverage has not obtained an institutional response to those accounts, and it does not present them as proof of current employer wrongdoing or a regulator’s legal liability.
SLPA’s maritime policy work can be tested against these practical questions: clear roles, verifiable qualifications, usable equipment, departure information, enforceable procedures and an auditable emergency response. Our draft remains an independent policy proposal. It is not enacted law or a text adopted by the Commission.
How to read and check this edition
All substantive testimony is paraphrased and linked to source passages. Day 2 part and paragraph references identify text in the supplied files; their timecodes are recording offsets, not collision times. Raw automated wording remains available, alongside clearly separated editorial notes. Day 1’s reviewed reading edition retains its originals and correction history. Unresolved names are not silently expanded and conflicting accounts are not harmonised.
Our assessment distinguishes personal observation, a statement read aloud, a questioner’s proposition, an inference and a procedural direction. None should be mistaken for a final finding. Search all six source parts, inspect the evidence register, or read the preserved Day 1 analysis.