x-pac

I’m a Ph.D scholar in Concordia University, Montreal, researching the computational nature and biological basis of Phonology as a cognitive ability. Here I’m supervised by Prof. Charles Reiss (Principal), Prof. Mark Hale and Prof. Alan Bale from Linguistics, as well as Prof. Roberto de Almeida from Psychology. My research is carried out under the umbrella of the INDI individualized doctorate program, in the form of an interdisciplinary investigation of the computational nature of our Language Faculty, specifically the sound domain, using both formal/mathematical and empirical cog-neuro tools. I’m supported by the Louise Dandurand Scholarship in Interdisciplinary Studies & the International Graduate Tuition Excellence Award. Prior to this I was a researcher in the MARCS Institute of Brain, Behavior and Development (Australia), and a PTF in The University of Auckland (New Zealand).

From time to time I also get fresh perspectives from Veno Volenec (Concordia), and Profs. Tom Bever (U. Arizona), Massimo Piatelli-Palmarini (U. Arizona), Martin Everaert (Utrecht), Bill Idsardi (UMD), Thomas Graff (Stony Brooke) and Norbert Hornstein (UMD). I strongly believe that the diversity of perspectives these experts provide leads to a very converging approach that leverages a wide variety of tools to understand the biological basis of our computational abilities.

I’m most interested in how generative and computational models of linguistic computation ought to address certain logical problems having to do with the mental representation of phonological knowledge, the nature and origin of phonological primitives, and how the phonological domain could interface with sensory and motor systems in the brain. My research is primarily motivated by the phonological theory promoted by Charles Reiss and Mark Hale, Substance-Free Phonology (Hale & Reiss, 2000, 2008), arguing that phonology is a computational system responsible solely for determining the combinatoric limits of the sound patterns of languages. As such, the primitives of phonology are computational primitives – symbolic entities devoid of any phonetic substance – and the principles of phonological computations are independent of functional phonetic pressures. Adopting a Fodorian modularity approach, SFP argues that the computational-representational phonological primitives and operations are opaque to sensory and motor concerns, even though there is direct and partially veridical link between the two. That is, while there indeed is a direct and predictable link between phonological primitives and their functional realizations, enabling a specific feature to always trigger, say, the rounding of the lips, this does not translate into that feature being reduced to the associated articulatory and perceptual specifics (i.e. SFP rejects any phonetic grounding). SFP argues that labels like [labial] or [coronal], a misleading nomenclature that explicitly or implicitly results in features being confused with their correlates, are a historical baggage for phonology. Instead, SFP argues that phonological features are merely algebraic variables, not unlike x or y, and that the ability to represent the concept of variables is genetically embedded in our species. The language-specific nature of phonetics implies that its specifics are not built-into this innate symbolic system. That specific features always trigger specific phonology-external sensory and motor correlates is achieved through first assigning specific values to variables, and then storing and recalling them from long term memory. Thus, the features themselves are mere algebraic variables, and only the computational combinatorics of phonology constrain their manipulation (i.e. the syntax of phonology is independent of functional pressures). The association of phonological primitives with their correlates is achieved through a transductive interface, the Phonology-Phonetics Interface (PPI). The theory of the PPI is termed Cognitive Phonetics (CP; Volenec & Reiss, 2018)). CP essentially serves the purpose of a linking hypothesis that seeks to explain the specifics of the PPI transducer. A brief overview of the argument would be something like this:

1. Modularity enforces legibility conditions on modules. Inner workings of a module is opaque from the outside.

2. Modules interface transductively, via an Input-Output (I-O) system where the output of one module undergoes transduction to be fed as the input to another module.

3. A transducer is a computational mechanism that translates the vocabulary of one module into that of another.

4. Both the input and the output of the phonological module are substance-free algebraic symbols, and are illegible to the sensory-motor systems.

5. The output of the phonological module, the surface representation (SR), is assigned specific sensory and motor correlates by the PPI.

6. The output of the PPI is the TPR, and it is directly legible by the sensory-motor systems.

The ability to store and recall such algebraic information is hypothesized to be the requirement for higher cognitive functions(Gallistel, 1981; Gallistel & King, 2009), and in recent years bioinformatics have supplied favourable evidence in support (Green et al., 2017). I am specifically interested in the computational architecture and the neurobiological implementation of PPI. Essentially a Marrist approach(Marr, 1982), my thesis is concerned with both formal elaboration of the architecture of PPI, and the specific neural machinery that sustains it. I approach the formal issues taking cue from the Chomskyan Biolinguistic program, while the neurobiological investigations are primary motivated by Granularity Mismatch (GM) and Ontological Incommensurability (OIP) arguments put forth by Embick, Poeppel and colleagues(Embick & Poeppel, 2015; Krakauer, Ghazanfar, Gomez-Marin, MacIver, & Poeppel, 2017; Poeppel & Embick, 2017). The central objective here is two-fold: (a) elaboration of the computational workings of Phonology, and (b) explaining the need for, and the nature of, a plausible interface theory compatible with our understanding of the functional and anatomical properties of the brain that support both thought and speech.


References:-

Embick, D., & Poeppel, D. (2015). Towards a computational (ist) neurobiology of language: correlational, integrated and explanatory neurolinguistics. Language, Cognition and Neuroscience, 30(4), 357–366.

Gallistel, C. R. (1981). Matters of principle: Hierarchies, representations, and action. Behavioral and Brain Sciences, 4(4), 639–650. https://doi.org/10.1017/S0140525X0000073X

Gallistel, C. R., & King, A. P. (2009). Memory and the Computational Brain. https://doi.org/10.1002/9781444310498

Green, A. A., Kim, J., Ma, D., Silver, P. A., Collins, J. J., & Yin, P. (2017). Complex cellular logic computation using ribocomputing devices. Nature, 548(7665), 117.

Hale, M., & Reiss, C. (2000). “Substance Abuse” and “Dysfunctionalism”: Current Trends in Phonology. Linguistic Inquiry, 31(1), 157–169. https://doi.org/10.1162/002438900554334

Hale, M., & Reiss, C. (2008). The phonological enterprise. Oxford University Press.

Krakauer, J. W., Ghazanfar, A. A., Gomez-Marin, A., MacIver, M. A., & Poeppel, D. (2017). Neuroscience Needs Behavior: Correcting a Reductionist Bias. Neuron, 93(3), 480–490. https://doi.org/10.1016/j.neuron.2016.12.041

Marr, D. (1982). Vision: A computational approach.

Poeppel, D., & Embick, D. (2017). Defining the relation between linguistics and neuroscience. In Twenty-first century psycholinguistics (pp. 103–118). Routledge.

Volenec, V., & Reiss, C. (2018). Cognitive Phonetics: The Transduction of Distinctive Features at the Phonology-Phonetics Interface (Vol. 11).