Word-production in aphasia: What impacts naming performance and recovery?

The ability to quickly and accurately produce a word is fundamental to human communication. In
neurologically healthy individuals, this feat is achieved through a complex interplay of several linguistic
processes. The details of these are not agreed upon by researchers, but must include at least: accessing
and selecting the word with the correct meaning (semantics); and selecting and assembling the correct
sounds (phonology).
Aphasia is a severe difficulty in understanding or producing language following a neurological injury. It
affects a large number of people as it occurs in about one third of people who suffer a stroke. Aphasia
severely limits verbal communication and social participation. Research shows that aphasia impacts
quality of life more severely than many serious health conditions including Alzheimer's Disease and
cancer.
Linguistic theory informs assessment and therapy in aphasia, which in turn refines our theoretical
understanding of language processing in the brain. People with aphasia find it difficult to produce the
words they intend to say. This is familiar to most of us as an occasional experience, when a word is on
"the tip of the tongue", but for people with aphasia the difficulty occurs for many common words, and
even in the apparently straight-forward task of single-word naming of objects or pictures, such as a cup
of tea. They may not be able to produce a word at all, or they may produce a word with semantic or
phonological errors. For example, in trying to say 'tea' an individual might instead say 'coffee' (a
semantic error) or 'key' (a phonological error). These errors can provide us with insight into the processes
which underlie naming in typical as well as disordered language. The mechanisms underlying such
naming impairments are, as yet, poorly understood and therefore inadequately treated. Clinicians profile
an individual's language with reference to models of word-production which do not fully account for the
data that we have, and the relationship between the type of impairment thus described and the
predicted success of a particular treatment approach is still poorly understood.

The aim of the proposed research is to refine linguistic models of word-production by using experimental
data from people with aphasia (PWA), taking an interdisciplinary and multimethod approach which uses
behavioural and computational data and psycholinguistic and cognitive science theories and frameworks.
In turn these models will be used to determine the best-suited intervention for an individual, based on
the type of language impairment which they have.
This aim will be addressed through three studies:
In Study 1, I will investigate the adequacy of different theoretical models to account for word production
deficits in aphasia, using existing open science datasets. Data on performance on naming tasks will be
analysed to investigate the influence of certain linguistic and image variables (such as length, frequency
and visual complexity) on the timing and accuracy of naming for PWA.
In Study 2, I will design and implement psycholinguistic experiments in aphasia to test the findings of
Study 1.
I will use a within-subject design to investigate how different types of cues which are commonly used in
clinical practice interact with linguistic and image variables. I will do this by carrying out trials of naming
tasks with PWA and analysing response times and error types.
In Study 3, I will develop and implement an intervention program for naming difficulties in aphasia by
incorporating the findings from the above two studies.
I will use a single-subject approach to test an intervention program for PWA based on the results from
Studies 1 and 2.