Bookshelf 📖

Articulatory Phonology: A phonology for public language use (Goldstein & Fowler., 2003) [READ pp. 1-15]

Citation

Goldstein, L., & Fowler, C. A. (2003). Articulatory phonology: A phonology for public language use. Phonetics and phonology in language comprehension and production: Differences and similarities, 159, 207.

Website: https://sites.google.com/view/ling580-2025/home

My thoughts

• 目标：建立一种把语言形式视为“公共行动”的理论，既能解释人们如何知道、产生、感知语言形式，又能解释这些形式如何在进化史与儿童发展中出现。
• 核心观点：音系应当描述可公开观察的言语行动的属性；生产与感知理论应分别解释这些行动如何被实现与被感知。由此预期音系形式会适应公共交流（说与听）的需求。

2.1 Phonology as a combinatoric system（音系的组合性）

• 人类语言用少量离散原子单位组合出大量词形，类似化学与遗传系统的“自多样化”结构，关键在于组合时原子单位保持离散身份。
• 实验记录（声学、发音、肌电等）显示连续与情境依赖，导致传统上把音段当作“心智”单位。
• AP主张：离散单位存在于更深层的“行动”层面，而不是表层的连续轨迹。

2.2 Units of combination are constriction actions of vocal organs（组合单位是发音器官的“收缩行动”）

三大假设：

• 单位是不同发音器官（唇、舌尖、舌身、舌根、腭帘、喉）的“收缩手势”（gestures），它们离散且相对独立，可作为组合“原子”。如 pack vs. tack 区分在于唇手势 vs. 舌尖手势。
• geature 通过“时间重叠的协调”形成更大结构（分子），重叠导致表面的同化/协同发音。用“任务动力学”通过相位关系建模 geature 协调与强度（bonding strength），强配对对应段内或韵首的紧密单元。
• geature以动力系统（如弹簧-质量系统）指定：在激活区间内参数固定，连续运动由系统方程决定，单位直接体现在运动学而非在心智“目标”与插值之间。
• 结果：可用“手势谱（gestural score）”展示多器官手势的时间安排与配对强度，层级结构（段、韵首/韵尾、音节、词足、词）可由配对强度涌现。

2.3 Evidence for gestures as units of speech production（证据）

• 分析-合成途径：从运动数据推断手势结构，再用模型生成匹配运动，得到支持。
• 口误/舌拙证据：在诸如“cop top”的快速重复中，常见“手势侵入”（intrusion）——如/k/的舌背手势侵入/t/位置，常为部分幅度并与目标手势并发，违反音系合法的线性段框架。
• 解释：把规划/生产单位视为手势，可将侵入视为系统自发转向更稳定的1:1协调模式；与英语既有的学习到的手势协调（不允许同时出现）竞争，产生侵入与偶发减弱。
• 要点：侵入是离散并发而非“混合中间姿态”；传统“抽象段错位”难以解释部分幅度与并发。

2.4 Phonological knowledge as constraints on gestural coordination（音系知识=对手势协调的抽象约束）

• 若形式是“协调的手势结构”，则音系普遍性应体现为对“协调模式”的约束，而非仅特征集合。
• 例1：闭塞序列的释放问题。Sierra Popoluca与摩洛哥阿拉伯语显示：同一抽象协调模式在同器官与异器官下产生不同释放表面，但语言的概括性必须诉诸“协调相位关系”，而非表面“有无释放”。该约束还与OCP等交互，影响模版填充和形态音变。
• 例2：音节结构的协调模式。韵首倾向“同步”（允许度内保证可辨识），韵尾倾向“顺序”。由此统一解释多种位置同位异音：
  • 元音在韵尾鼻音前鼻化：腭帘降低手势与口腔收缩在韵尾呈先后关系，叠加致鼻化；韵首则同步不致鼻化。
  • [l] 的清亮/暗化：舌尖闭合与舌背回缩的相对时序在韵首/韵尾不同，导致感知上的[-back]/[+back]差异。
  • 类似差异亦见于[w]与[r]。

Parity in public language use（公共语言使用中的“同型性/对等性”）

• 生产与感知需要“共同货币”（common currency）：语言知觉、产出与知识中的单位应一致（言者与听者共享同一形式）。
• 来自更广泛的感知-行动领域的动机：例如新生儿跨模态模仿现象显示需要超越特定感觉/运动编码的“共同编码”，信息应指向“远端对象与事件”（如舌头与其前伸动作），而非仅近端刺激。只有共享的公共行动单位（手势）才能支持感知指导行动与交流中的“同型性”（parity）。

总体结论

• 发音音系学将音系单位界定为“发音器官的收缩手势”，通过动力学与相位协调形成更大结构，解释了表面连续、情境依赖的发音现象与口误数据，并把音系知识表述为对协调模式的抽象约束。该框架满足交流中的“共同货币/对等性”需求，统一生产与感知。

Theoretical goals

1. develop a realistic understanding og language forms as language users know them, produce them, and perceive them
2. Aim to understand how the forms might have emerged in the evolutionary history of humans and how they arise developmentally, as a child interacts with speakers in the environment

Concepts

- Language forms (public actions)

• entities of various grain sizes that theories of phonology characterize, which means languages provide to make between-person linguistic communication possible, not the mental categories of most theories of phonology, production, and perception

- A theory of phonology

• Theory about the properties of those public actions

- A theory of speech production

• How those actions are achieved

- A theory of speech perception

• How those actions are perceived

- A theory of the emergence of phonological structure in language

• How particulate language forms emerged in the course of communicative exchanges between people
• predicts the forms will have properties that adapt them for public language use
  • for speaking
  • for perception
  • from acoustic speech signals

Articulatory phonology

1. Phonology as a combinatoric system

   • Speech is a continuous, context-dependent motion of a large number of degrees of freedom; the actions themselves that engage the vocal tract and regulate the motions of its articulators are discrete and context-independent

   • phonological units are abstract in being coarse-grained (low-dimensional) with specific motions of articulators, and to the acoustic structure that may specify the motions

     • Most fundamental property of speech communication -> phonological structure
   • small (<100) inventory of primitive units to combine in different ways -> form words (vocabularies of human languages)
   • combinatoric property -> chemical compounding and genetic recombination

• self-diversifying systems
  • atomic units are distinct from one another
  • retain discreteness when they combine to form new objects
  • The diversity of combinations tends to decrease as more and more atoms join
  • toward the mean value of the scalar units

  • recurring substructures

  • hierarchy

    • show that combination involves: creation of new objects when atoms retain their discrete identities
  • segments = phonological units

1. Units of combination (atoms) are constriction actions of vocal organs
   • phonological units have dual roles
2. units of action
3. units of combination (and contrast)

   Three key hypotheses

4. The vocal tract activity can be analyzed into constriction actions of distinct vocal organs
5. Actions are organized into temporally overlapping structures
6. Constriction formation is appropriately modeled by dynamical systems

2.2.1. Constriction actions and the organs that produce them

6 distinct organs

• lips
• tongue tip
• tongue body
• tongue root
• velum

• larynx

• gestures = constricting actions of distinct organs
  • atoms of a combinatoric system = satisfy the property of discrete differences
• 2 combinations of geatures
  • can be defined as potentially contrasting with one another
  • if : include at least one distinct constriction geatures
  • example: pack vs. tack (pack includes a lips geature, taks includes a tongue tip geature)

2.2.2. Coordination of gestures and overlap Articulatory phonology hypotehsiszes

• geatures are coordinated into more elaborate molecular structures in which geatures can overlap in time
• coproduction = account for superficial context-dependence that is observed in speech

• onsets = ions of a combinatoric system: internally cohesive structures of atoms that recombine readily with other such structures

2.2.3. Dynamical specification

• geatural units are dynamical systems

2.3. Evidence for gestures as units of speech production

• speech errors = more direct envidence for decomposing speech prodcution and/or planning into geatural strucutres
• segments are the most common units involved in errors (changes in position, such as anticipations, perseverations and exhanges)

2.4. Phonological knowledge as (abstract) constraints on gestural coordination

• for native speakers’ implicit knowledge of phonological structure and regularities in a particular language
  • inventory of lexically contrastive units
  • constraints on phonological forms
  • systematic alternations to lexical forms that result from morphonological combination and embedding in a particular prosodic context
  • phonological generalizations = patterns or modes of coordination
    • two closute geatures will determine the resulting aerodynamic and acoustic consequenese
    • syllable strcutre = disctinct modes of geatura coordination for consonant geatures iin an onset vs. coda
    • Onset = synchronous mode of coordination dominantes
      • consonant geatures
    • coda = sequential mode dominates
      • synchronous prodcution is most compatible with recoverability when a narrow construction (lips, tonegue tip) is coproduced with a wider one (laryngeal or velic geature, voiceless or nasal stops, secondary palatalizaiton, verlarizaiton or rounding)

1. Parity in public language use
   • the language forms that the language users know, produce and perceive must be the same

3.1. The need for a common currency in perceptually guided action, including speech

• need for a common currency, the transmitted and received messages may be the same, is known as the parity requirement
• perception and action
  • example: imitation by infants, newborns are disposed to imitate the facial geatures of an adult