ERC 10th Anniversary Week: How does the brain process language?

Language and Brain Lab Team

 

Aditi Lahiri and her research team have been fortunate to receive ERC funding for 2 projects in a row, all centring around the theme of words.

ERC Advanced Grant 2011-16 - Words: Asymmetry, change and processing in phonological mental representation

This project investigated the temporal and historical dimension of mental representations of words. Although speech is extremely variable such that no word is ever pronounced the same way twice, the human brain is able to understand and comprehend words without any obvious difficulty. The project pursued four research questions, where the answers to each called for a dialogue between a variety of disciplines:

  1. what is the nature and phonological structure of mental representations of words and how are they constrained? 
  2. how are these representations processed and accessed in the course of everyday communication?
  3. how and why do representations change, while sometimes tenaciously remaining constant over time?
  4. can the hypotheses and predictions about mental representations be computationally modelled?

The research included synchronic and diachronic linguistic analyses, and combines techniques from historical linguistics (studying original manuscripts, developing rules of sound change), behavioural psycholinguistic experiments (with on-line lexical decision tasks), brain-imaging experiments (using EEG) and computational modelling.

The project presented here was challenging. No single model of word representation, processing and change had ever been put forward to integrate seemingly diverse processes of language comprehension, language change and variation. Nevertheless, to have a comprehensive idea of how the human processing system works and how it changes, we must have separate pieces of evidence from different research areas. If we succeed to tie together the theoretical analyses and experimental evidence, testing the insights with computational modelling, it would indeed be a major break-through. 

Proof of Concept - 2014

This research has resulted in a ‘Proof of Concept’ grant, which has converted a research tool of the  ‘Words’ project into a working single-word Automatic Speech Recognition (ASR) demonstration system to attract speech technology companies to use this method in commercial ASR systems. This planned single-word ASR system [the Flexible Speech Recognition system (FlexSR)] has evolved from a linguistic model of phonological representation, which is the focus of the present ERC grant.

Two patents have now been submitted and the team are currently working on their business plans to take this tool further.

ERC Grant 2016 – 21 - Resolving Morpho-Phonological Alternation: Historical, Neurolinguistic, and Computational approaches

The opaque phonological relationship between morphologically related forms has been a long-standing challenge in theoretical, historical, psycho- and neuro-linguistics, and computational linguistics alike. Morpho-phonological alternations of all kinds have been analysed across the languages of the world; but fundamental questions have remained controversial or indeed unasked:

  • Why do they exist in the first place and why are they so widespread?
  • How do they come about and what is their diachronic time-course?
  • How are they represented in mental lexicons and how are they processed?

Rather than setting morpho-phonological alternations aside as irregularities of morphology (requiring individual listing and storing), we recognise certain kinds of them (stress shifts, feature changes, deletions, and tonal changes) as something universally to be expected in mental lexicons and as something the brains of speakers and listeners can easily handle. The position that we advocate is that morpho-phonological variants are not listed and stored independently, but rather are mapped onto single abstract representations. This is a controversial position, and its defence requires the systematic study of types of alternations and their histories, and precise hypotheses about the nature of mental representations.

 

The ERC

The ERC grant has made a massive different to Aditi Lahiri, not just as it has meant that she has been able to research a topic, that is exactly and specifically based on her area of expertise, but to the wider humanities division as well. It has given the Linguistics Faculty a chance to increase the number of Early Career Researchers that are working around the subject. This has given those ECRs excellent experience and has led to many going onto bigger things following their time on the project.

The freedom that the grant offers when applying for support for a research project is amazing and many of the discoveries and advancements that this project has made would certainly not have been possible without the support of the ERC.

List of site pages