Chapter 45: The Question She Couldn't Answer
Celeste answered the committee's question by narrowing her title.
"Founder Representative was not a scientific approval role," she said. "I had no authority to approve participant-risk content."
No one disputed the second sentence.
Her use of Founder Representative was already part of the old meeting record. The issue was not whether the title existed. It was what she had chosen to do while using it.
The restricted responsibility review began with the two version codes on a single screen. Leah's documents remained sealed. Only their permitted control fields appeared.
The chair repeated the question.
"Who authorized release of the second final version?"
Celeste looked toward Nathan.
He answered as counsel. "The current record does not identify a scientific approver for that specific version."
"Then who authorized its distribution?"
"Scientific approval and distribution are different functions," Celeste said.
"Agreed. Answer the distribution question."
She returned to her lack of scientific authority.
I submitted one record from Sterling's own release-routing production. It contained no patient name, health information, or document text. The committee had authorized it because it addressed timing and company contact.
The email came from Celeste's account two days before the later explanation packet was sent. Her signature read FOUNDER REPRESENTATIVE.
Move the participant explanation to release today. We cannot lose another investor communications window to an internal review cycle.
Sterling's own archive authenticated the sender, date, and routing fields. The message had been produced as part of the company-contact record. I did not need Leah's private file to prove Celeste had written it.
The email could not identify an author, editor, or scientific approver. Its authenticated sender and timing proved something else: Celeste had pressed for release before another internal review cycle finished.
"That is scheduling," Celeste said. "Not scientific approval."
"Yes. And scheduling was the power you used."
"Investor communications window referred to the timing of a general company update," she said. "It was not an instruction to change participant material."
"The email does not accuse you of changing the material," I said. "It records you pushing it out before review finished."
She had expected me to reach past the evidence. My agreement left her with the part the email actually proved.
"You told the committee your lack of scientific authority removed you from the decision," I said. "It did not remove you from the pressure applied to the people making it."
The chair entered the distinction into the finding. Celeste had participated in release timing and invoked a founder-level role. The email did not establish who altered any document or approved its scientific content.
Nathan requested that the email remain restricted. Leah's attorney agreed. Consent to review company contact did not authorize publicity, and the committee had no need to publish the text to preserve its finding.
Celeste still had not named the person who approved the second final version.
The chair asked Sterling for the contemporaneous distribution record behind its index. Nathan produced the current archive entry. It identified the later packet but contained no original authentication showing which version had been exported for release.
"Does Sterling have a contemporaneous export or authentication record?" the chair asked.
Nathan conferred with outside counsel.
"Not in the current archive."
That answer was narrower than no.
"Was there no record at the time," the chair asked, "or is there no record in the archive you searched?"
"I can confirm only the archive before us."
The committee ordered Sterling to search its authorized historical systems and produce any contemporaneous record capable of authenticating the released version. The order did not name a format, a storage location, or a person. It required only verifiable provenance.
Celeste requested a week. The chair gave Sterling until nine the next morning because the company had already represented its production as complete.
Nathan reserved Sterling's objections and accepted the deadline.
Nine the next morning.
The screen went dark, leaving the two final-version codes on the printed agenda between us.
By morning, "not in the current archive" would no longer be an answer.