A digitised archive claimed that a 1924 hypnosis（催眠）/hɪp'nəʊsɪs/ trial had caused a coma（[医] 昏迷）/'kəʊmə/. The claim arose because optical recognition converted every label on a blank infirmary（医务室）/ɪn'fɜːm(ə)rɪ/ checklist into a recorded event. Before further handling（在进一步接触处理之前）/bɪˈfɔː ˈfɜːðə ˈhændlɪŋ/, data historian Eli Ren sealed the scans and asked a clinical therapist（临床医学家）/'θerəpɪst/ to open an inquiry（启动核查）/ˈəʊpən ən ɪnˈkwaɪəri/. Nothing in the file（档案中没有任何材料）/ˈnʌθɪŋ ɪn ðə faɪl/ yet showed that any dangerous state had occurred. In light of（鉴于）/ɪn laɪt əv/ that absence, the reviewers began with the original paper rather than the searchable transcript.

The sleep vocabulary supplied the first boundary. A sleepless（失眠的）/'sliːplɪs/ volunteer might drowse（打瞌睡）/draʊz/, become drowsy（昏昏欲睡的）/'draʊzɪ/, and later enter slumber（睡眠）/'slʌmbə/, while dormant（休眠的）/'dɔːm(ə)nt/ described unused apparatus. Stupor（昏迷）/'stjuːpə/ and coma were clinical states, not synonyms for ordinary sleep. Awareness（意识）/ə'weənəs/ and consciousness（意识）/'kɒnʃəsnɪs/ were observation headings; subconscious（潜意识的）/sʌb'kɒnʃəs/ named a theoretical category. Awaken（唤醒）/ə'weɪk(ə)n/ and waken（唤醒）/'weɪk(ə)n/ appeared as spoken instructions, and lucid（明晰的）/'luːsɪd/ described a clear response. At the alleged moment（在所称的那个时刻）/æt ði əˈledʒd ˈməʊmənt/ of collapse, the technician had written that the volunteer answered promptly.

The protocol then separated procedure from outcome. Hypnosis and hypnotism（催眠术）/'hɪpnətɪz(ə)m/ named the method under study; hypnotic（催眠的）/hɪp'nɒtɪk/ modified the recorded suggestion. The therapist would mesmerise（施以催眠术）/'mezməraiz, 'mes-/ no one by force but would present a verbal stimulus（刺激）/'stɪmjʊləs/. Volunteers had to memorise（（英）记忆）/'meməraɪz/ a short list, daydream（做白日梦）/'deɪdriːm/ about a journey, use imagination（[心理] 想象力）/ɪ,mædʒɪ'neɪʃ(ə)n/, and envision（想象）/ɪnˈvɪʒn/ an imaginative（富于想象的）/ɪ'mædʒɪnətɪv/ scene. One schema（[计][心理] 模式）/'skiːmə/ grouped these tasks for analysis. Minutes before（在……前几分钟）/ˈmɪnɪts bɪˈfɔː/ the supposed emergency, the same volunteer had completed every task and requested water.

A mood questionnaire explained another cluster of false events. Bemused（困惑的）/bɪ'mjuːzd/, emotional（情绪的）/ɪ'məʊʃ(ə)n(ə)l/, egocentric（[心理] 自我中心的）/,egə(ʊ)'sentrɪk; iː-/, melancholy（忧郁的）/'melənkəlɪ/, and disillusioned（使醒悟）/,dɪsɪ'l(j)uːʒ(ə)n/ were answer options. Euphoria（精神欢快）/juː'fɔːrɪə/, the doldrums（忧郁）/'dɒldrəmz/, funk（恐惧）/fʌŋk/, and the tendency to mope（忧郁）/məʊp/ or act as a pessimist（悲观主义者）/'pesimist/ formed separate scales. Frightening（令人恐惧的）/'fraɪt(ə)nɪŋ; 'fraɪtnɪŋ/ and horrify（使恐惧）/'hɒrɪfaɪ/ evaluated the script, not the participant. Hallucination（幻觉）/hə,luːsɪ'neɪʃ(ə)n/ was a prompt, while hallucinate（使产生幻觉）/hə'luːsɪneɪt/ was its verb. Disenchant（使清醒）/dɪsɪn'tʃɑːnt; dɪsen-/ and disillusion described changes in audience expectation. The period heading psycho（心理分析）/'saɪkəʊ/ was editorial shorthand, and lunacy（精神失常）/'luːnəsɪ/ was an obsolete label; neither was a modern diagnosis.

Not until Eli compared the marks with the printed form did the physical findings become clear. The checklist named eyelid（[解剖] 眼睑）/'aɪlɪd/ and eyelash（睫毛）/'aɪlæʃ/ movement, saliva（唾液）/sə'laɪvə/ and salivary（唾液的）/'sælɪvərɪ/ flow, and any tendency to drool（流口水）/druːl/. It also listed an involuntary（无意识的）/ɪn'vɒlənt(ə)rɪ/ twitch（抽搐）/twɪtʃ/, spasm（[临床] 痉挛）/'spæz(ə)m/, tremor（[医] 震颤）/'tremə/, flaccid（[医] 弛缓的）/'flæsɪd; 'flæksɪd/ muscle, paralysis（麻痹）/pə'rælɪsɪs/, and a paralytic（麻痹的）/pærə'lɪtɪk/ response. Ventricular（心室的）/ven'trɪkjʊlə/ trouble, asthma（[内科][中医] 哮喘）/'æsmə/, and apoplexy（[医] 中风）/'æpəpleksɪ/ were exclusion conditions; stupefaction（麻醉）/,stʃuːpɪ'fækʃn; ,stjuːpɪ'fækʃn/ was a historical heading for anaesthesia. Every box beside those terms was empty. The software had read the labels but discarded the marks.

Eli told the archive board, “A checklist can name a risk, but it cannot prove that the risk occurred.” At this stage（现阶段）/æt ðɪs steɪdʒ/, the board withdrew the automated conclusion. Archivists rebuilt the transcript in accordance with（按照）/ɪn əˈkɔːdəns wɪð/ the visible marks and linked each observation to its page image. The board's directive was simple: make the data available（公开提供数据）/meɪk ðə ˈdeɪtə əˈveɪləbəl/. Reviewers could then inspect the distinction themselves. Beyond that（除此之外）/bɪˈjɒnd ðæt/, the archive made no claim about whether the old experiment was therapeutically useful. It reported only that the alleged coma had never been recorded.