Substance-free Phonology

The concept of "inner speech" is fundamental to cognitive science and Brain-Computer Interface (BCI) research, yet it often conflates two distinct processes: the auditory experience of a "voice in the head" (subvocalization) and the mental simulation of muscle movements for speech (silent articulation). This paper proposes a theoretical framework to dissociate these phenomena from a computational linguistics perspective. We posit Substance-Free Phonology (SFP) as a theoretical prior to argue that subvocalization corresponds to computation at the abstract, symbolic level of phonology, while silent articulation involves the subsequent mapping of these symbols onto a concrete, substance-based phonetic motor plan. This distinction yields specific, testable neurophysiological predictions and provides a more nuanced foundation for developing next-generation, languagebased BCIs.

We investigated whether silent articulation (covert speech) produces alpha-band event-related desynchronization (ERD) patterns analogous to the well-established ERD/ERS signatures of motor imagery, and whether such signatures can be detected on a single-trial basis. Healthy participants (N = 6) performed an EEG experiment with four conditions: covert articulation of real words (CAW), overt articulation of real words (OAW), covert articulation of pronounceable nonce words (CNW), and overt articulation of nonce words (ONW), plus a noarticulation control. EEG was recorded with a 64-channel EGI NetStation system; analyses focused on electrodes covering the peri-central and superior temporal regions. After preprocessing, extended-infomax ICA was used to remove ocular and lateral artefacts, data were re-referenced to average, and time-frequency decomposition (Morlet wavelets) was used to quantify alpha (8-12 Hz) power. Group-level time-frequency maps showed a reliable alpha power decrease time-locked to condition cues for both covert and overt articulation. On a single-trial level, an ERD detection rule (positive percent change from baseline to active window) yielded a detection success rate of roughly 70-75% across conditions (i.e., per-trial error rate ≃ 25-30%). We describe the experimental design, signal-processing pipeline, and statistical analysis (cluster permutation tests for ERPs and nonparametric tests for single-trial ERD accuracy), and we discuss implications for silent-speech brain-computer interfaces and for models of speech planning. Our results show that silent articulation produces measurable alpha ERD but highlight limitations in single-trial reliability that must be considered for online decoding.

Los términos fortis-lenis se han empleado para nombrar diferentes tipos de contrastes consonánticos en las lenguas zapotecas. En este trabajo argumento que estos térmonos hacen referencia a contrastes relacionados históricamente entre sí, pero independientes en la sincronía. De la misma forma, se elabora una propuesta de desarrollo histórico de los sistemas consonánticos de las lenguas zapotecas y se reconstruyen tres procesos fonológicos que existieron en el pasado y ya no están activos en la actualidad: sonorización de obstruyentes, develarización y asimilación total al inicio.
También hago un recuento de la terminología empleada en los estudios de lenguas zapotecas para nombrar los contrastes de fonación. Argumento que en el zapoteco de San Bartolo Yautepec, la voz laringizada es un rasgo prosódico que interactúa con el tono.

Review of Iosad, Pavel, A Substance-Free Framework for Phonology: An analysis of
the Breton dialect of Bothoa, Edinburgh Studies in Theoretical Linguistics 2,
Edinburgh: Edinburgh University Press, 2017, x + 274 pp. Hardback ISBN
978 1 4744 0737 3: £75 (reduced £52.50); webready pdf ISBN 978 1 4744
0738 0: £75 (reduced £52.50); epub ISBN 978 1 4744 0739 7: £75 (reduced
£52.50).

This talk explores the nature of the phonology-phonetics interface in light of current neurobiological evidence, the legacy of the Sound Patterns of English, and the re-emerging Analysis-by-Synthesis program in Biolinguistics to argue for a partially veridical framework where feature-substrates and expectations, but not features (qua features), are innate. We argue that the innate and universal component of phonological grammars contain only (a) the neural substrates that support the implementation and computation of distinctive features, and (b) an innate, but not language-specific, ability to create features based on experience with acoustic and articulatory phenomena. Like Chomsky & Halle (1999: pp.19) we argue that there exist no conditions of linearity and invariance between phonological and phonetic representations, and that post-phonological spell-out functions exactly the same way as spell-out at the upper interface. In particular, we point out that the framework of "Cognitive Phonetics" is, both, empirically untenable and conceptually incoherent.

Rules built from a single operation, priority union, can model both feature-filling and feature-changing processes. The distinction is handled by specifying which argument of priority union is defeasible. Priority union should be considered as a candidate for inclusion in phonological Universal Grammar.

We propose that the interface between phonology and phonetics is mediated by a transduction process that converts elementary units of phonological computation, features, into temporally coordinated neuromuscular patterns, called 'True Phonetic Representations', which are directly interpretable by the motor system of speech production. Our view of the interface is constrained by substance-free generative phonological assumptions and by insights gained from psycholinguistic and phonetic models of speech production. To distinguish transduction of abstract phonological units into planned neuromuscular patterns from the biomechanics of speech production usually associated with physiological phonetics, we have termed this interface theory 'Cognitive Phonetics' (CP). The inner workings of CP are described in terms of Marr's (1982/2010) tri-level approach, which we used to construct a linking hypothesis relating formal phonology to neurobiological activity. Potential neurobiological correlates supporting various parts of CP are presented. We also argue that CP augments the study of certain phonetic phenomena, most notably coarticulation, and suggest that some phenomena usually considered phonological (e.g., naturalness and gradience) receive better explanations within CP.

The article illustrates some of the salient features of Government Phonology (GP) 2.0 by axiomatising (a subclass of) the set of possible Putonghua forms. We show that a phonological theory can profit by assuming that phonological representations are hierarchical, just like syntactic representations. A structural relation of c++command, a relative of the well-known c-command, is used heavily. The similarity with syntax is further underlined by the introduction of a phonological Binding Theory: illicit representations are prohibited by the LUxI Principles, the phonological counterpart of the Principles A, B and C.

It is fashionable these days to decry the assumption that the primes of phonological computation could be innate (Mielke, 2004), or that their interpretation at the interface with phonetics could be universal and deterministic (Dresher, 2009). The arguments against universality and determinism, both, ultimately reduce to a few related observations stemming from the nature of the relationship between the primes over which phonology computes, and the way said primes relate to actual attested speech forms (Mielke, 2004; B. D. Samuels, 2011). It is argued, for instance, that different segments bearing identical specifications for a feature, say {-Round} in /i/ vs. /e/, externalize it differently, with no invariant substantive correlates. This lack of invariance is alleged to be a concern for universal/innate features. Likewise, in languages like Malayalam, where velar consonants are palatalized in the context of the vowels /i, e, a/, it is often assumed that the low back vowel /a/ bears a {-Back} specification even though phonetically it is, in fact, a back vowel (Mohanan & Mohanan, 1984; Warrier, 1976). Samuels (2011) and Mielke (2004), therefore, argue that since some features (e.g. the {-Back} of Malayalam /a/) are to be determined based purely on phonological activity, it makes little sense to have any pre-specified features at all. In this now predominant view of primes (

This paper investigates the nature of the phonetics-phonology interface-the relationship between phonetic substance and phonological primes-in light of a pattern of velar palatalization in Malayalam. Data from psycholinguistic investigations of the process are interpreted in the context of a modular framework with a transductive interface between phonological symbols and phonetic substance. It is argued that the modular, interfacedriven account provides for a more economical, and biolinguistically coherent, account than substantive and grounded approaches.

We propose that the interface between phonology and phonetics is mediated by a transduction process that converts elementary units of phonological computation, features, into temporally coordinated neuromuscular patterns, called ‘True Phonetic Representations’, which are directly interpretable by the motor system of speech production. Our view of the interface is constrained by substance-free generative phonological assumptions and by insights gained from psycholinguistic and phonetic models of speech production. To distinguish transduction of abstract phonological units into planned neuromuscular patterns from the biomechanics of speech production usually associated with physiological phonetics, we have termed this interface theory ‘Cognitive Phonetics’ (CP). The inner workings of CP are described in terms of Marr’s (1982/2010) tri-level approach, which we used to construct a linking hypothesis relating formal phonology to neurobiological activity. Potential neurobiological correla...

This study addresses a puzzle in Croatian regarding the fate of the underlying palatal glide in intervocalic position. Approaching the puzzle from the perspective of Cognitive Phonetics (CP), we advance two claims: First, output phonological representations consisting of features are not directly interpretable by the articulatory system; rather, the interface between the phonology and the articulatory system is mediated by a universal transduction system. Second, the main units of speech production are transduced phonological features, and not segments, syllables, or articulatory gestures.

Slides from the course entitled The Logical Structure of Phonological Theory given at the Eastern Generative Grammar summer school 2022 in Brno, Czech Republic. The course provides an introduction to a substance-free model of phonological knowledge called Logical Phonology. The main tenet of that approach is that phonological knowledge can be described and partly explained using simple operations characteristic of logical set-theory, such as set unification, intersection and subtraction. Logical Phonology will be applied to data from a variety of languages to illustrate the main principles of this approach, its analytical methods, and its limitations.

Rules built from a single (partial) operation, priority union, can model both featurefilling and feature-changing processes. The distinction is handled by specifying which argument of priority union is defeasible. Priority union should be considered as a candidate for inclusion in phonological Universal Grammar.

Resumen: En una de las variedades del ese eja (Takana), existe un fenómeno de desnasalización por el cual las consonantes nasales de esta lengua se realizan como nasales simples, nasales complejas (con grados de nasalidad variable) y oclusivas sonoras simples. En esta investigación se propone un análisis fonológico de dicho fenómeno con base en evidencia acústica. Los datos analizados fueron obtenidos de entrevistas llevadas a cabo con hablantes de ese eja de la comunidad de Palma Real (Tambopata, Madre de Dios). La descripción acústica de la desnasalización permitió identificar y caracterizar el conjunto de variantes de las consonantes nasales, su frecuencia y las restricciones lingüísticas sobre su ocurrencia. En el análisis fonológico, enmarcado en la teoría de optimalidad (en inglés, Optimality Theory) y el modelo de estructuras paralelas (en inglés, Parallel Structures Model), se demuestra y formaliza que la desnasalización en ese eja es un proceso de oralización de consonantes nasales originado por las vocales orales inmediatamente siguientes. Los resultados de esta investigación contribuyen (i) al debate del papel del contraste entre vocales orales y nasales en la ocurrencia de la oralización de consonantes nasales, o protección (en inglés, shielding), al demostrar que es posible que una lengua que carece de nasalidad vocálica contrastiva manifieste dicho proceso; y (ii) a la discusión de la representación fonológica de la actividad del velo en el modelo de estructuras paralelas.
Palabras clave: ese eja, Takana, desnasalización, protección, segmentos de contorno nasales, teoría de optimalidad, modelo de estructuras paralelas.

Rules built from a single operation, priority union, can model both feature-filling and feature-changing processes. The distinction is handled by specifying which argument of priority union is defeasible. Priority union should be considered as a candidate for inclusion in phonological Universal Grammar.

Rules built from a single  operation, priority union, can model  both feature-filling and feature-changing processes. The distinction is handled by specifying which argument of priority union is defeasible.  Priority union should be considered as a candidate for inclusion in phonological Universal Grammar. 
Keywords: {phonology, feature logic, rules, Universal Grammar}}

This study addresses a puzzle in Croatian regarding the fate of the underlying palatal glide in intervocalic position. Approaching the puzzle from the perspective of Cognitive Phonetics (CP), we advance two claims: First, output phonological representations consisting of features are not directly interpretable by the articulatory system; rather, the interface between the phonology and the articulatory system is mediated by a universal transduction system. Second, the main units of speech production are transduced phonological features, and not segments, syllables, or articulatory gestures.

U izlaganju se predstavlja kognitivna fonetika, neurobiološki utemeljena teorija o sučelju između fonologije i fonetike. Osnovna je njezina tvrdnja da se izlazni oblici iz fonološke sastavnice gramatike pretvaraju u vremenski koordinirane živčano-mišićne upute; ta se pretvorba odvija s pomoću dvaju algoritama – paradigmatskoga pretvorbenoga algoritma koji svakomu razlikovnomu obilježju dodjeljuje živčano-mišićnu aktivnost i sintagmatskoga pretvorbenoga algoritma koji tu aktivnost rasprostire u pravome vremenu. Prikazuju se određeni empirijski rezultati kognitivne fonetike, posebice na području proučavanja koartikulacije. Tom se teorijom premošćuje konceptualni jaz između suvremene generativne fonologije i postojećih psiholingvističkih i neurolingvističkih modela jezične uporabe, a u širemu kontekstu kognitivne neuroznanosti kognitivna je fonetika neurobiološki osviješten doprinos razumijevanju odnosa između jezične sposobnosti i jezične uporabe.

This paper provides an analysis of Croatian nasal place assimilation from the perspective of Substance Free Logical Phonology. Treating phonology as application of logico-mathematical functions to abstract, symbolic primitives such as features, the analysis explicitly captures both the fact that in Croatian each underlying nasal behaves differently with respect to place assimilation and the fact that in certain cases nasals assimilate in continuancy as well as in place. On the basis of the provided analysis and on theoretical grounds, it is argued that not all possible why-questions that could be asked about phonology are grammatically relevant: True phonological why-questions, those that receive explanation within a formal grammar, should be strictly distinguished from functionalist why-questions, which concern themselves with aspects of phonology that are reducible to language-external systems and should not be encoded in a generative grammar.

We propose that the interface between phonology and phonetics is mediated by a transduction process that converts elementary units of phonological computation, features, into temporally coordinated neuromuscular patterns, called ‘True Phonetic Representations’, which are directly interpretable by the motor system of speech production. Our view of the interface is constrained by substance-free generative phonological assumptions and by insights gained from psycholinguistic and phonetic models of speech production. To distinguish transduction of abstract phonological units into planned neuromuscular patterns from the biomechanics of speech production usually associated with physiological phonetics, we have termed this interface theory ‘Cognitive Phonetics’ (CP). The inner workings of CP are described in terms of Marr’s (1982/2010) tri-level approach, which we used to construct a linking hypothesis relating formal phonology to neurobiological activity. Potential neurobiological correlates supporting various parts of CP are presented. We also argue that CP augments the study of certain phonetic phenomena, most notably coarticulation, and suggest that some phenomena usually considered phonological (e.g., naturalness and gradience) receive better explanations within CP.

There is widespread agreement that both phonological knowledge and spoken linguistic signals  manifest putative universals, which, systematically used, enable infinite productivity. However, less agreement prevails on whether the phonological system is combinatorial  or blending or in nature. In a combinatorial system (DNA is an example), discrete units are combined recursively to yield infinite forms, while maintaining the individuality of each block. For linguistic forms, optimal transmission generally assumes that some contextual blending of phonemes is necessary. It has been variously claimed that transmissibility thus renders phonology merely a summary requirement imposed by the speech channel, while others argue that such purely blending systems cannot account for highly unnatural yet productive patterns that have been attested, and that the greedy reductionism of cognitive abilities to motor functions amounts to substance abuse. An alternative view holds that while phonology is indeed algebraically combinatorial, it is also  subject to functional pressures albeit only selectively. We here report a test of the latter alternative hypothesis.  Our data clearly illustrate that while phonological processes are, in fact, affected by channel concerns, these are attended to selectively in a manner that does not override grammatical requirements. In light of the distinct but converging findings from the two experiments, we conclude that phonology manifests algebraically combinatorial rules that are selectively subject to functional optimization in order to attain discrete infinity while maintaining transmissibility.