审阅范围 · Review scope
指定内容与证据核对。请保留正文中的出处与限制；不代表全库最终验收、实际演奏或学习效果验证。
Specified content and evidence checks. Keep the evidence and limitations stated in the text in view. This is not final corpus acceptance or validation of performance or learning outcomes.
内听与结构记忆：先形成预期，再核对
先在心里想一小句音乐，再弹出来，看看哪些地方与预期相同。例如把 D4–F4–E4–A4 各弹一拍：先记下你预计它会怎样上行、下行，再用实际声音核对。这里检查的是自己的预期，不要求第一次就猜准。
在没有实际发声时想象音乐，本文称为“内听”。Chang 的 Mental Play（心中演奏）还包含他对离琴记忆、动作和谱面预想的教学安排，不能只凭这个词断定所有人脑中的体验相同。所需前置知识是音名与拍；“四分音符”在下面的自编练习中按每个音一拍处理。
【作者观点】Chang 在 FOPP-15 把 Mental Play 描述为在心中演奏音乐，并用于离琴检查记忆；FOPP-45 把看谱、视唱/内听与回到琴上核对联系起来。本轮读取这两节英文提取稿，并在实际 PDF 正文确认相关段落。§15 起于 PDF 实际第38页（印刷37），§45 起于实际102页（印刷101）；没有据此把其余未读章节算入覆盖。
分开检查“记得”什么
【本库综合】一段音乐可以留下不同线索：手的移动顺序、谱面位置、听觉轮廓、节奏组织或和声路径。熟练地从头接着弹，可能掩盖其中某些线索尚不清楚。
【本库综合】在离琴时先预想一个很短的自编句子，记下你能确定的节奏、轮廓或音名，再到琴上核对。若卡住，记录具体位置；不要把“想起来的感觉很熟”直接算成可独立重现。
连接处也是记忆的一部分
【作者观点】FOPP-09 将可同时构成配置的音组与连接音分开讨论，提醒局部速度练习还需要接回循环与前后文。来源为 PDF 实际第22–25页相关正文（节首印刷21）。作者也明确说这种练习并不等于古典作品的最终演奏。
【本库综合】将一个片段压缩成和弦或骨架，会暂时移除时间信息。恢复原谱时，需要重新检查节奏、声部、重音与连接；骨架会弹不能证明这些层次已经恢复。
保留哪些边界
【证据边界】原著对年龄、绝对音感、演出焦虑和广泛学习效果有很强的判断，本轮未独立验证。本库的实验不把绝对音感设为准入条件，也不承诺某种内听练习必然消除忘谱或焦虑。§9 的数学速度类比和脑功能解释也没有被当作科学结论采纳。
用实验 E11区分预期与实际结果，再与练习方法、听觉训练相互检查。
记忆不必只靠从头接着弹
【作者观点】新读的 FOPP-14 把键盘位置、动作习惯、声音和结构分析都纳入记忆讨论，并建议不同位置起弹。该节同时采用先背后练的强烈取向；本库未把“人人必须如此”或照片式记忆、年龄、智力的保证采纳为事实。
【本库综合】选择一个可独立说清的入口，例如“第三个音后上跳，再回到起点”，用谱面核对。能够描述结构、能够从入口发声和能够连续演奏是不同结果。保留看谱核对的途径，见当场、保留与迁移。
联想提示与生理解释要分开
【作者观点】新读的FOPP-16 用熟悉结构作为回忆提示，并明确把“记忆场”的说明称为自己的假说。实际PDF第40–43页相关正文与中文第1–4、7段已比较；两处“假说”限定保留了下来，但这不验证该模型。
【本库综合】可以为短句留下音程走向、重复位置和和弦的文字提示，再测试是否帮助找回入口。这是可记录的任务；它不证明所有经历都会永久存储，也不证明量子类比解释了记忆。既有中文对savant添加了诊断式限定，另记于纠错日志，本库不据此作人群推断。
从陌生记谱建立声音预期
【作者观点】FOPP-45实际103–104页（印刷102–103）建议先读未听过的材料、唱或在心中演奏，再回琴核对。这支持“先预测、后检查”的任务来源；作者关于绝对音感和必然进步的强断言仍未验证。
【通用乐理】固定唱名把 do 绑定音名C；首调唱名在大调中把 do 作为主音。选择哪套标签与是否拥有绝对音感是不同问题。Open Music Theory 的相关定义已核对；本例只用大调，不混入小调唱名的不同体系。
【本库综合·原创例子】先给C大调及参考音C4，读 C4 D4 E4；另给G大调及参考音G4，读 G4 A4 B4。两个音列的相邻距离都是2、2半音，首调唱名都是 do–re–mi；固定唱名分别是 do–re–mi 与 sol–la–si（此处用si命名B）。音名、调内关系和唱出的准确性分别检查。得到参考音后能唱出关系，不代表无需参考即可识别音名；本例的字母音列也不是已测过的五线谱视唱能力。
【本库综合】循序检查：先预测起音与休止，再给参考音并预想一小段的高低走向，愿意时唱或哼出来，最后用琴声或已核对的声音答案比较。先记预测再听答案；反复播放之后的熟悉不能冒充第一次从谱建立声音。下一轮换未听过、难度可控的短谱；记录节奏、轮廓、音程、调内关系和与参考音的音高是否各自吻合。能口头算出音程与能在心中听见它也不同；唱不准可能还包含发声困难，不能自动诊断内听错误。逐渐加入声部与和声，但一条旋律的成功不证明能内听整首多声部音乐。见练习项目PP10与反馈PP11。
先记录，再听目标
打开原创双语内听图例：先看一小节4/4，按需听参照C4，把轮廓与时值预测写下，再主动揭示／播放。A为 C4 D4 E4 C4 各1拍；B为 C4 E4 D4，时值1、1、2拍。这段文字已经透露答案，因此读过后再试同谱要标记为复习。新题测试需要真正未听过的材料。
页面记录是否打开文字谱面、是否听参照以及是否已揭示同谱；它不从文字答案自动推断内听能力。声音是正弦合成音，不是钢琴采样；浏览器生成的样本频率和时长已检查，但本任务未直接听辨扬声器输出。完整课程仍是待验证研究，不是由一次点击证明有效。
不把案例译意变成记忆机制
【翻译复核】FOPP-16第2段（实页40／印页39）中的案例按上下文是事故及紧邻其前几分钟记不起；译稿一处却写成只记得那几分钟。这是缺失区间与保留区间的反转。
【证据边界】Q-060只纠正作者案例的文字关系，不验证统一5分钟转存或永久记忆机制，也不提供健康判断。练习仍用自己的预想、弹奏与延迟复测记录，不能由这段案例推出固定的背谱等待时间。
分清同时与依次，再比较结果
【翻译复核】FOPP-09第12段（实23／印22）在上下文中区分三音依次起音与临时和弦简化，译稿却增加“同时”。一次手部动作并不决定有几次起音；Q-068修复这个条件，既不规定落手幅度，也不要求初学者快速复现原练法。
【本库综合】原创纸上例：把C4、E4、G4全部写在第1拍同一列，是同时起音；把它们分别写在第1、2、3拍，则是依次起音。两者音高集合相同，起音时刻不同。先遮住答案，说清自己预测的是哪一版；此表没有指定时值、指法或作品和声功能，不能凭音名集合补出这些信息。答案分别为1个共同起音时刻与3个先后起音时刻；不是说第一版只含1个音。
【证据边界】同节第19段（实24／印23）是作者的潜在伤害警示；译文把它写成确定的风险排序。Q-072不提供发生率或个人安全评估，也不把次日没有酸痛当作自动增加练习量的许可。
先认出改变了什么，再决定能复用多少练习
【作者观点】FOPP-67把重复动机与数学变换联系起来。找相同和不同可作为分析入口，但本章对莫扎特统一公式、贝多芬群论意图、性别／语言联想和观众心理的解释，仍是作者观点。本批没有核完相关作品的谱本和录音，不能从相似形状推断作曲家的意识或学习成效（Q-083／Q-084；实174–181／印173–180）。
【本库综合】原创对照表：每行都是单旋律的1小节4/4；每拍是四分音符，时值列以拍为单位。先不弹，指出哪一列改变。C4是中央C；数字是八度标记，F#4表示该八度的升F。这里不指定指法、奏法或和声功能。
| 版本 | 依次音高 | 依次时值（拍） | 与原型比较 |
|---|---|---|---|
| 原型 | C4 E4 D4 C4 | 1,1,1,1 | 4次起音，合4拍 |
| 全部升高2半音 | D4 F#4 E4 D4 | 1,1,1,1 | 音程路线相同，实际键不同 |
| 音高顺序倒读 | C4 D4 E4 C4 | 1,1,1,1 | 时间顺序改变，时值不变 |
| 只改时值 | C4 E4 D4 C4 | 0.5,0.5,1,2 | 起音在第1、1.5、2、3拍；仍合4拍 |
自查：每行总长相同，但最后一行前两个音各半拍，末音两拍；同一拍号不等于相同节奏。把音程／节奏识别列为一个小任务，把实际键位、指法可行性和前后连接列为另一个。可在纸上完成前者，不把它算成已会弹后者；下次遮住比较栏重新辨认，再接回实际短句。表中已给答案，不能当成未见材料的迁移测验。
【通用定义与本库模型】群需要封闭的组合运算、结合律、单位元和逆元；参见MIT定义1.1，第3页。可选的原创数学模型只用0–11表示十二个音高类，C=0：Tn(x)=(x+n) mod 12。T0对每一类都不改；先T2再T10会返回原音高类。这里的群元素是这12个变换，运算是接着做另一个变换，不是把四个相同音直接叫单位元。模型忽略八度与拼写，回到同一类不保证回到同一键、同一功能或原指法；它也不证明作品是晶体空间群。
【来源纠错】FOPP-14第12段（实33–34／印32–33）把十进制与模9混同。十进制用0–9；取非负整数末位对应模10，原例5621除以10余1、除以9余5。Q-079保留这项原书错误的归属；知道八度循环也不能代替全部88键的定位与连接检查。
FOPP-14第40段及FOPP-82书目中的Sinfonia，在本书§52第26段明确指巴赫三部创意曲，非管弦乐交响曲（Q-080）。正确曲类名称帮助找对材料，但不要求初学者立即练整首。
Inner hearing and structural memory: predict, then check
Imagine a tiny phrase, then play it and compare. For D4–F4–E4–A4, one beat per note, first record the expected rises and falls; then check the actual sound. This tests your prediction, without requiring a correct first attempt.
Imagining music without sounding it is called inner hearing here. Chang's Mental Play also includes his teaching arrangements for memory, movement and score visualization away from the piano. The term does not establish identical mental experiences for everyone. Prerequisites: note names and beat. In the original exercise below, each quarter note receives one beat.
[Author's view] FOPP-15 describes Mental Play as playing music in the mind and uses it to check memory away from the piano. FOPP-45 connects notation, singing/inner hearing and keyboard checks. Both English extracts were read and relevant actual PDF passages checked. Section 15 begins on physical p. 38 (printed 37); section 45 on physical p. 102 (printed 101). Other unread sections are not included by association.
Separate what is remembered
[Our synthesis] Memory can involve movement sequence, score position, auditory contour, rhythm or harmonic route. Fluently continuing from the beginning can conceal gaps in some of these cues.
[Our synthesis] Away from the piano, imagine a very short original phrase. Record the rhythm, contour or names you can specify, then check at the keyboard. Locate gaps instead of equating familiarity with independent reproduction.
Connections also belong to memory
[Author's view] FOPP-09 distinguishes simultaneously playable note groups from conjunctions, reconnecting local speed work to cycles and context. Relevant physical pp. 22–25 were checked; its opening printed page is 21. The author also says this exercise is not the final form of classical performance.
[Our synthesis] Compressing a passage into a chord or skeleton temporarily removes temporal information. On restoring the score, recheck rhythm, voices, accents and connections. Playing the skeleton does not establish these other layers.
Limits retained
[Evidence limits] Strong claims about age, absolute pitch, performance anxiety and broad learning outcomes have not been independently verified here. Our exercise neither requires absolute pitch nor promises to prevent memory failures or anxiety. Section 9's mathematical speed analogy and brain-function explanation are also not adopted as scientific conclusions.
Use E11 to separate prediction from result, alongside practice methods and ear training.
Memory need not depend only on continuing from the beginning
[Author's view] Newly read FOPP-14 discusses keyboard location, movement habits, sound and structural analysis, recommending different entry points. It also strongly favors memorizing before practice. Universal requirements and guarantees about photographic memory, age or intelligence are not adopted as facts here.
[Our synthesis] Choose an independently describable entry, such as a leap after the third note followed by return to the start, and check the score. Describing structure, sounding an entry and playing continuously are separate results. Retain notation checks; see immediate performance, retention and transfer.
Separate associative cues from physiological explanations
[Author's view] Newly read FOPP-16 uses familiar structure as a recall cue and explicitly calls its memory-field account the author's hypothesis. Relevant physical PDF pages 40–43 and Chinese paragraphs 1–4/7 were compared. Both hypothesis qualifications remain, without validating the model.
[Our synthesis] Leave written cues for interval direction, repetitions and harmony, then test whether they help retrieve an entry. This is an observable task, not proof that all experiences are permanently stored or that a quantum analogy explains memory. The existing Chinese adds a diagnostic restriction to savant; the correction log records it without inferring group identities here.
Build sound expectations from unfamiliar notation
[Author's view] On physical pages 103–104 (printed 102–103), FOPP-45 recommends reading unheard material, singing or mentally playing it, then checking at the piano. This supports the source of the prediction-before-checking task; strong claims about absolute pitch and inevitable progress remain unverified.
[General theory] Fixed do binds do to the note name C; movable do in major uses do for the tonic. A labeling system and possession of absolute pitch are different questions. The relevant Open Music Theory definitions were checked; this example uses major only, without mixing minor-solfege systems.
[Our synthesis · original example] Give C major and reference C4, then read C4 D4 E4; separately give G major and reference G4, then read G4 A4 B4. Both have successive distances of 2, 2 semitones and movable-do syllables do–re–mi. Fixed syllables are do–re–mi and sol–la–si respectively, using si for B here. Check names, tonal relationships and singing accuracy separately. Singing relationships after a reference does not establish reference-free pitch naming; these letter-name sequences also do not constitute a tested staff-sight-singing ability.
[Our synthesis] Check progressively: predict onsets and rests, then take a reference pitch and imagine a short contour, optionally sing or hum it, and compare with piano sound or a verified sonic answer. Record the prediction before hearing the answer; familiarity after repeated playback cannot stand in for first-time notation-to-sound prediction. Next, try an unheard, manageable short score, recording rhythm, contour, intervals, tonal relationships and pitch against the reference separately. Calculating an interval verbally differs from internally hearing it; inaccurate singing can also involve vocal production, so it does not automatically diagnose faulty inner hearing. Add voices and harmony progressively, but one successful melody does not establish whole-piece polyphonic inner hearing. See practice-project PP10 and feedback PP11.
Record first, hear the target later
Open the original bilingual audiation illustration: read one 4/4 bar, optionally hear reference C4, write a contour/duration prediction, then explicitly reveal/play. A is C4 D4 E4 C4, one beat each; B is C4 E4 D4, lasting 1, 1 and 2 beats. This paragraph has disclosed the answers, so a later attempt with the same material is review. A transfer test requires genuinely unheard material.
The page records text-notation aid, reference sound and prior revelation of the same pattern. It does not infer audiation from a written answer. Sound is synthesized sine tone, not a piano sample. Frequencies and durations of browser-rendered samples were checked, but this task did not directly audition speaker output. The full curriculum remains a research proposal, not an efficacy result established by a click.
Do not turn a translated anecdote into a memory mechanism
[Translation review] The example in FOPP-16 paragraph 2 (physical 40/printed 39) concerns missing memories of the incident and the immediately preceding minutes. One translated sentence instead describes remembering only those minutes, reversing the missing and retained intervals.
[Evidence limits] Q-060 corrects only the relationship in the author’s example. It validates neither a uniform 5-minute transfer nor permanent memory and provides no health judgment. Practice still uses your prediction, playing and delayed-retest records; the anecdote does not establish a fixed waiting time for memorization.
Separate simultaneous and successive conditions before comparing results
[Translation review] In context, FOPP-09 paragraph12 (physical23/printed22) distinguishes successive three-note attacks from temporary chord simplification; the translation adds “simultaneous.” One hand movement does not determine the number of attacks. Q-068 restores this condition without specifying a drop height or requiring beginners to reproduce the fast original drill.
[Our synthesis] Original paper example: place C4, E4, G4 in one column on beat1 for simultaneous onsets; place them on beats1,2,3 respectively for successive onsets. The pitch collection is the same but onset times differ. Hide the answer and identify which version you predict. No durations, fingering or harmonic function in a piece are specified, so a note-name set cannot supply them. Answers:1 shared onset time versus3 successive onset times, not only1 note in the first version.
[Evidence limits] Paragraph19 (physical24/printed23) gives the author’s potential-harm warning, translated as a definite risk ranking. Q-072 supplies neither incidence nor individual safety assessment and does not treat absence of next-day soreness as automatic permission to increase practice.
Identify the change before deciding how much practice transfers
[Author's view] FOPP-67 links recurring motifs with mathematical transformations. Similarity/difference can provide an analytical entry point, but its unified Mozart formula, Beethoven group-theory intentions, gender/language associations and audience-psychology explanations remain author views. This batch did not verify the relevant complete scores/recordings; similar shapes do not establish compositional consciousness or learning effects (Q-083/Q-084; physical174–181/printed173–180).
[Our synthesis] Original comparison table: each row is one monophonic bar of4/4 with a quarter-note beat; durations are in beats. Before playing, identify the changed column. C4 is middle C, digits mark octaves, and F#4 is F sharp in that octave. Fingering, articulation and harmonic function are unspecified.
| Version | Successive pitches | Successive durations (beats) | Comparison with original |
|---|---|---|---|
| Original | C4 E4 D4 C4 | 1,1,1,1 | 4 onsets totaling4 beats |
| All pitches up2 semitones | D4 F#4 E4 D4 | 1,1,1,1 | Same interval route, different keys |
| Pitch order reversed | C4 D4 E4 C4 | 1,1,1,1 | Temporal ordering changes; durations do not |
| Durations only | C4 E4 D4 C4 | 0.5,0.5,1,2 | Onsets at beats1,1.5,2,3; still4 beats total |
Self-check: all rows have the same length, but the last row starts with two half-beat notes and ends with a two-beat note. Identical meter does not mean identical rhythm. Make interval/rhythm recognition one task and actual keys, feasible fingering and joins another. Paper work can complete the first without establishing the second. Next time hide the comparison column, identify again, then reintegrate into the real phrase. Answers are disclosed here, so this is not an unseen-material transfer test.
[General definition and our model] A group requires closed composition, associativity, identity and inverses; see MIT definition1.1, page3. An optional original model uses0–11 for twelve pitch classes, C=0: Tn(x)=(x+n) mod 12. T0 changes no class; T2 followed by T10 returns the original class. The group elements are these12 transformations and the operation is successive composition, not labeling four repeated notes an identity. Register and spelling are omitted: returning to a class does not guarantee the same key, function or fingering. Nor does this prove a piece is a crystallographic space group.
[Source correction] FOPP-14 paragraph12 (physical33–34/printed32–33) conflates decimal notation and modulo9. Decimal notation uses0–9; a nonnegative integer’s final digit is its remainder modulo10. The source example5621 leaves1 modulo10 and5 modulo9. Q-079 attributes this to the original book; knowing octave cycles cannot replace location/join checks across all88 keys.
Sinfonia in FOPP-14 paragraph40 and the FOPP-82 bibliography is explicitly a Bach three-part invention in section52 paragraph26, not an orchestral symphony (Q-080). Correct repertoire labels support finding appropriate material without requiring a beginner to play the entire piece immediately.